The growing awareness in health care of the challenges associated with proper dosing of anticoagulants such as heparin has inspired the emerging field of anticoagulation stewardship.
Anticoagulants are an increasingly common cause of adverse drug events (ADEs). Recent studies show that anticoagulants surpassed antibiotics as the primary source of emergency department (ED) visits for ADEs in the United States nearly a decade ago. Data through 2019 show anticoagulants were responsible for 21.5% of ED visits for ADEs, with antibiotics a distant third at 12.8%.1
ADEs related to anticoagulants, such as unfractionated heparin, low-molecular weight heparin, and warfarin include excessive bleeding and thrombosis, both of which can be fatal. A 5-year study at Brigham and Women’s Hospital (BWH) in Boston showed that 11% of patients with anticoagulation-associated ADE died within 30 days.2 Elderly and cardiac patients, in particular, are at high risk for anticoagulant-related ADEs.
The BWH study also showed a median length of stay of 13 days for patients who suffered ADEs and a total mean hospitalization cost per ADE of $58,991, a near doubling from the $29,851 stated cost prior to the ADE, with most post-ADE expenditures attributable to nursing and pharmacy costs.
The growing awareness in health care of the challenges associated with proper dosing of anticoagulants such as heparin has inspired the emerging field of anticoagulation stewardship, spearheaded by the Anticoagulation Forum. The goal of Anticoagulation Stewardship Programs is to develop initiatives to optimize anticoagulation-related health outcomes and minimize avoidable ADEs.3
Though anticoagulation stewardship is in its early stages compared with antimicrobial stewardship—a field whose roots go back to the mid-1990s)—it is gaining momentum. Anticoagulation stewardship is actively supported by the US Department of Health and Human Services,4 the FDA,5 and the National Quality Forum.6 And at least 8 major US hospitals and medical centers have implemented anticoagulation stewardship programs.3
Heparin is used to prevent or inhibit the development and growth of blood clots in patients with certain medical conditions or patients undergoing high-risk medical procedures. Unfortunately, despite advances in therapeutic heparin dosing and monitoring, it is difficult to determine the right dosage for a specific patient because individuals process heparin at their own rate based on body size, blood composition, and underlying disease state. Standardized dosing guidelines fail a proportion of patients and this introduces risk.
When heparin levels are below the target range for a patient, the result may be worsening clot formation, with potential for embolic organ failure or embolic stroke and death. Should heparin levels exceed the target range, major bleeding may occur, potentially triggering hemorrhagic stroke, organ failure, and death. Thus, it is critical for clinical staff to closely monitor heparin’s effect on a patient.
The 2 most common measures of heparin’s activity are the anti-factor Xa assay (anti-Xa) and activated partial thromboplastin time (aPTT) lab tests. Each of these present problems because the logistics of sampling heparin values at precisely the right time create challenges for nurses, whose schedules are already overburdened. If sample draws are timed incorrectly, they must be discarded, which wastes hospital resources. Worse, poorly timed heparin tests could be misinterpreted, possibly resulting in misguided dosing decisions that increase patient risk.
Model-informed precision dosing (MIPD) has proven clinically beneficial in achieving therapeutic drug exposure and decreasing the need for dosing adjustment with multiple drugs. MIPD has been shown to work particularly well with vancomycin management.7 There has been little research, however, into the application of MIPD to the dosing of unfractionated heparin.
But there are similarities between heparin and vancomycin that indicate MIPD holds promise in heparin use cases. Both vancomycin and heparin have:
Further, both heparin and vancomycin are used on a diverse patient population, have no widely accepted “standard” dosing for all patients, and have validated means to measure exposure. However, there are also some distinct challenges associated with heparin that software-based or MIPD-based solutions will have to consider.
Promisingly, model-based methods are well-suited for handling both these problems. For example, a model could consider the action of multiple drugs on an endpoint such as anti-Xa, facilitating the switch from long-acting outpatient medications to fast-acting inpatient medications. A model could also describe the impact of one or more anti-coagulation drugs on more than 1 endpoint, helping clinicians find a regimen that is likely to achieve therapeutic levels of both anti-Xa and aPTT.
Recently, scientists at InsightRX conducted a simulation to determine whether MIPD could be applied to heparin dosing and yield similar results to vancomycin. Specifically, researchers sought to determine whether a MIPD strategy for initial dosing would achieve therapeutic anti-Xa levels more frequently than a hospital dosing nomogram.
Simulation results indicate that using MIPD to select an initial dose is more likely to bring a patient to target drug exposure levels than a typical hospital dosing nomogram. The research team’s findings were presented at the 17th National Conference on Anticoagulation Therapy in April as a poster titled “Anti-Xa target attainment & dose selection: comparison of pharmacokinetic model- guided dosing versus institutional practice.”
InsightRX employed a digital twin study with an anti-Xa target range of 0.5 to 0.7 IU/mL. The study used a sample of 2000 real adult patients from a precision dosing database to provide real-world covariate distributions.
Initial doses were then calculated using either a typical hospital nomogram or using population PK model parameters to select a dose that would be expected to lead to anti-Xa levels of 0.5-0.7 IU/mL. For the nomogram, both a “high dose” and a “low dose” nomogram were implemented, reflecting different clinical indications.
For the PK model approach, 2 different pharmacokinetic/pharmacodynamic (PKPD) models were used: the Delavenne model and the Brunet model.8,9 Next, patient response to these doses was simulated by randomly sampling from the inter-individual variability distribution described by both PKPD models.
Initial dosing using the Delavenne model was in the target range 54.7% of the time, while initial dosing based on the Brunet model hit the target range 15.8% of the time. Both models had higher rates of successfully bringing patients to target exposure levels than the low-dose (11.4%) or high-dose (0.0%) hospital nomogram without requiring further dose adjustments.
The simulation results indicate that initial heparin dosing using MIPD is more likely to achieve target steady-state anti-Xa values than a hospital nomogram. Initial heparin dosing regimens using a hospital nomogram are likely to provide supratherapeutic exposure.
This finding is significant because more accurate initial heparin dosing will reduce the number of patients whose doses must be adjusted and will minimize the likelihood of an ADE that could be life-threatening or life-diminishing to a patient. InsightRX plans to validate results of this simulation using real-world data.
It is clear that therapeutic monitoring is needed to optimize patient outcomes in heparin therapy. While further study of hospital practice and clinical data is warranted to identify the role of MIPD in heparin therapy, the InsightRX simulation suggests that the model-informed precision dosing software and workflow that has been used successfully for vancomycin can be applied to heparin dosing.
This will benefit patients and overworked hospital staff and may lead to a reduction in dangerous and costly anticoagulation-associated ADEs. MIPD soon may do for heparin what it has done for vancomycin.
Finally, InsightRX’s research into MIPD and heparin dosing could take the anticoagulation stewardship movement to the next level by providing tools and a dosing model that can be used to minimize ADEs and drive better outcomes for patients treated with heparin rather than doses being calculated via hospital nomograms.
About the Authors
Sirj Goswami, PhD, is CEO and co-founder of InsightRX.
Jon Faldasz, PharmD, BCPS, is senior director Product and Customer Experience for InsightRX.
Jasmine Hughes, PhD, is director of Data Science for InsightRX.
1. Burnett A,Rudd K, Triller D. Advancing anticoagulation stewardship: A call to action for stewardship from the US-based anticoagulation forum, Thrombosis Update. Science Direct. Volume 9, 2022, 100125, ISSN 2666-5727. https://doi.org/10.1016/j.tru.2022.100125.
2. Piazza G, Nguyen TN, Cios D, Labreche M, Hohlfelder B, Fanikos J, Fiumara K, Goldhaber SZ. Anticoagulation-associated adverse drug events. Am J Med. 2011 Dec;124(12):1136-42. doi: 10.1016/j.amjmed.2011.06.009. PMID: 22114827; PMCID: PMC3224344. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3224344/
3. Education and Guidance. Anticoagulation Forum. Web page. Available at https://acforum.org/web/education-stewardship.php. Accessed August 14, 2023.
4.National action plan for ADE prevention. Available at: https://health.gov/about-odphp/previous-initiatives/national-ade-action-plan. August 14, 2023. https://health.gov/sites/default/files/2019-09/ADE-Action-Plan-Anticoagulants.pdf
5. U.S. FDA. Core Elements of Anticoagulation Stewardship Program. Available at: https://cacmap.fda.gov/media/132114/download. Accessed August 14, 2023.
6. National Quality Forum (NQF) Releases Anticoagulation Stewardship Playbook to Help Improve Patient Safety. National Quality Forum. News release. August 24, 2022. Accessed August 14, 2023. https://www.qualityforum.org/News_And_Resources/Press_Releases/2022/National_Quality_Forum_(NQF)_Releases_Anticoagulation_Stewardship_Playbook_to_Help_Improve_Patient_Safety.aspx
7. Frymoyer A, Schwenk HT, Zorn Y, Bio L, Moss JD, Chasmawala B, Faulkenberry J, Goswami S, Keizer RJ, Ghaskari S. Model-Informed Precision Dosing of Vancomycin in Hospitalized Children: Implementation and Adoption at an Academic Children's Hospital. Front Pharmacol. 2020 Apr 29;11:551. doi: 10.3389/fphar.2020.00551. PMID: 32411000; PMCID: PMC7201037.
8. Delavenne X, Ollier E, Chollet S, Sandri F, Lanoiselée J, Hodin S, Montmartin A, Fuzellier JF, Mismetti P, Gergelé L. Pharmacokinetic/pharmacodynamic model for unfractionated heparin dosing during cardiopulmonary bypass. Br J Anaesth. 2017 May 1;118(5):705-712. doi: 10.1093/bja/aex044. PMID: 28510738.
9. Brunet P, Simon N, Opris A, Faure V, Lorec-Penet AM, Portugal H, Dussol B, Berland Y. Pharmacodynamics of unfractionated heparin during and after a hemodialysis session. Am J Kidney Dis. 2008 May;51(5):789-95. doi: 10.1053/j.ajkd.2007.12.040. PMID: 18436089.