Highlights
- •Limited data exist on use of anticoagulants in newborns, children, and adolescents.
- •Healthcare claims data show infrequent use of anticoagulants in children.
- •Heparins were used most frequently, followed by warfarin.
- •Direct oral anticoagulants were rarely used.
- •Less than half of children with history of Fontan procedure received anticoagulants
ABSTRACT
Purpose
Children generally have a lower risk of venous thromboembolism (VTE) than adults, but those with acute and chronic conditions requiring hospitalization and surgical procedures are at increased risk. Anticoagulant use in children has not been systematically studied, and limited data exist. This study aimed to provide data on the conditions associated with use of anticoagulants, the type of anticoagulant used in children, and the incidence of thromboembolism and major bleeding events reported in this population.
Methods
To increase understanding of the use of anticoagulant therapies in children with at-risk conditions, 3 health claims databases in the United States were analyzed to describe the characteristics of use of heparins, warfarin, and direct oral anticoagulants (DOACs). Cumulative drug exposure was determined for continuous exposure, defined as >30 days. Unadjusted event rates of VTE and major bleeding after exposure to these therapies were reported. The data were presented descriptively and are not intended for comparison or to imply any causation.
Findings
Anticoagulants were infrequently used in the pediatric population, including at any time point after Fontan surgery for congenital heart disease. Heparins were used most frequently in the population overall and especially for patients aged <12 years. DOACs were used least often and primarily for patients ages 12 to <18 years. Among pediatric patients exposed to anticoagulants, unadjusted incidence rates of VTE per 1000 person-years of exposure ranged from 30.8 to 34.0 for all DOACs, 21.6 to 46.2 for warfarin, and 6.0 to 7.3 for heparins. Rates per 1000 person-years for major bleeding ranged from 0 to 4.9 for all DOACs, 4.3 to 6.7 for warfarin, and 3.7 to 4.6 for heparins.
Implications
With results from clinical trials evaluating DOACs in the pediatric population expected in the next 2 years, these descriptive real-world data may provide a baseline understanding of current prescribing patterns and outcomes associated with the use of DOACs and other anticoagulants in routine pediatric clinical practice. This information represents the use of real-world evidence and may function as the benchmark for evaluating changes in prescription practices and potential outcomes in the future.
Keywords
Introduction
The incidence of venous thromboembolism (VTE) in children is very low, at 0.07 to 0.14 per 10,000 children but increases 100 to 1000 times among hospitalized children to ≥58 per 10,000 admissions.
1
Among pediatric patients with VTE, >90% have more than one risk factor, such as a central venous access device, infection, cancer, and/or congenital heart disease (CHD).2
,3
Pediatric VTE is a severe problem with the potential for significant complications or death, and recurrent VTE has been reported to occur in up to 15% of children after an initial event.3
Thrombosis is a frequent complication in ill neonates, in children with CHD after a Fontan procedure, and in adolescents with associated risk factors. Central venous catheters, fluid fluctuations, sepsis, liver dysfunction, and inflammation contribute to the risk profile for VTE development.
4
Mechanical ventilation, perinatal asphyxia, CHD, and dehydration are also recognized risk factors for neonatal VTE. The Fontan procedure is the final palliative surgery performed in children with CHD with single-ventricle physiology, which affects ∼10% of children with CHD.5
,6
The Fontan procedure is the most common procedure performed for these children after the age of 2 years. Thrombosis reportedly occurs in 17% to 33% of patients from 2 days to >10 years after Fontan surgery.7
, 8
, 9
The risk of thrombosis is highest within 6 months of the procedure, after which it diminishes but persists, with a gradual increase over the first 2.5 years.10
,11
In adolescents, VTE is frequently associated with risk factors such as malignancies, use of central venous catheters, immobilization, and use of oral contraceptives.Currently, there is limited approval of anticoagulant therapies in the pediatric population. Anticoagulant therapy may be used in pediatric patients who are at increased risk of thrombotic events, but there is no consensus in the literature or routine clinical practice about the optimal type or duration of these therapies. Furthermore, the associated outcomes from these therapies are not well described in the literature. Pediatric clinicians often rely on data from adult patients. In general, the American Society of Hematology recommends using anticoagulation rather than not using anticoagulation in pediatric patients with symptomatic thrombotic events.
1
For children having the Fontan procedure, antithrombotic guidelines in the United States recommend the use of acetylsalicylic acid (ASA) or unfractionated heparin (UFH) followed by a vitamin K antagonist.12
, 13
, 14
However, these guidelines are based on limited data, which are largely extrapolated from adult populations based on pharmacologic mechanism, and are variably followed in clinical practice.UFH is commonly used in pediatric patients but is not without challenges.
3
Heparin binds to antithrombin and inactivates coagulation factors IIa, Xa, XIa, and XIIa.15
There are age-related differences in levels of antithrombin and prothrombin, activity of anti–factor Xa, and plasma protein binding. The effects of these differences on heparin efficacy and safety have not been studied in pediatric patients, and target therapeutic ranges are therefore extrapolated from adults. Achieving target activated partial thromboplastin time or anti–factor Xa levels often leads to frequent blood draws in neonates and children, which are complicated by low blood volumes and low circulating levels of antithrombin.16
,17
These young patients often require antithrombin infusions to optimize the anticoagulant effect of heparin. Low-molecular-weight heparin also binds to antithrombin but requires no routine monitoring of anticoagulant activity and can be administered subcutaneously.15
However, reduced levels of antithrombin in neonates require age-specific dosing of low-molecular-weight heparin. Warfarin inhibits vitamin K, decreasing coagulation factors II, VII, IX, and X and proteins C and S. In addition to frequent laboratory monitoring, the use of warfarin in infants requires the highest per kilogram dose, which can be difficult to administer due to the lack of a liquid oral formulation.15
In pediatric patients undergoing extracorporeal membrane oxygenation, the direct thrombin inhibitor bivalirudin is a commonly used alternative to UFH despite limited dosing information.18
The direct oral anticoagulants (DOACs) are more appealing for use in pediatric patients based on their direct effects on factor Xa or thrombin, their stable pharmacokinetic profile, and lack of need for routine monitoring. No DOACs are currently approved in the United States for use in children; however, two Phase III studies have recently been completed.
2
,19
There are also currently >20 clinical trials of DOACs in children and adolescents registered at ClinicalTrials.gov, with VTE as the primary efficacy objective of the investigation. Because DOACs have established a favorable benefit–risk profile compared with warfarin and other standard of care in adult settings, it is expected that this new generation of drugs may offer alternative therapeutic options in the pediatric population, although this is pending the results of further clinical investigations.Additional studies including real-world data are needed to evaluate the current use of anticoagulant therapy in children with at-risk conditions and to document future practice changes. Using 3 large US health claims databases, the present descriptive study reports the baseline conditions potentially associated with use, including medical conditions and drug exposures, for pediatric populations treated with DOACs and other anticoagulant therapies. In addition, the unadjusted event rates of VTE and major bleeding after exposure to anticoagulants in the pediatric population, and more specifically in pediatric patients who have had a Fontan procedure, are presented.
Patients and Methods
Study Design and Data Sources
This retrospective analysis identified patients <18 years of age from the following US databases over the period January 1, 2010, to December 31, 2019: IBM MarketScan Commercial Claims and Encounters Database (CCAE), IBM MarketScan Multi-State Medicaid Database (MDCD), and Optum's de-identified Clinformatics Data Mart Database–Socio-Economic Status (Optum). The databases were transformed into the Observational Medical Outcomes Partnership (ie, the OMOP common data model).
20
,21
CCAE includes adjudicated health insurance claims for employer-sponsored health plans, including fee-for-service, preferred provider organizations, and capitated health plans, and provides inpatient, outpatient, outpatient pharmacy, laboratory tests, and enrollment information. MDCD includes adjudicated health insurance claims for Medicaid enrollees in 10 to 12 states and provides hospital discharge diagnoses and outpatient diagnoses, procedures, and pharmacy claims. Optum is an adjudicated administrative health care claims database for members with private health insurance, including primary US commercial claims for patients ages 0 to 65 years, with some Medicare claims (age ≥65 years). Data captured in Optum include inpatient and outpatient medical services, prescriptions as dispensed, and outpatient laboratory tests processed by large national vendors who participate in data exchange with Optum.
The use of the CCAE, MDCD, and Optum databases was reviewed by the New England Institutional Review Board and was determined to be exempt from broad Institutional Review Board approval as this project did not involve human patient research.
Assessments
These analyses considered the following variables based on the lists of codes provided in Supplemental Tables I to IV (see the online version at doi:10.1016/j.clinthera.2021.09.021). Age was categorized into 5 groups: <1, 1 to <2, 2 to <7, 7 to <12, and 12 to <18 years. Sex was also recorded. Drugs of interest were defined at the ingredient level and included DOACs (ie, rivaroxaban, apixaban, edoxaban, dabigatran), warfarin, and heparins (ie, UFH, enoxaparin, nadroparin, dalteparin). Cumulative drug exposure was determined for continuous exposure defined as >30 days, and gap days longer than 30 days were not included in the cumulative exposure. Mean drug exposure was calculated. Patients could be counted in more than one drug exposure category.
To describe clinical conditions that might be associated with use of anticoagulant therapies, the following conditions were identified in the 183 days prior to and including the drug exposure start date for the pediatric population: cerebral vein and sinus thrombosis (CVST); VTE, including deep vein thrombosis and pulmonary embolism; and fractures. From a clinical perspective, these conditions may be viewed as heightened risks for thrombotic events due to underlying diseases or restricted mobility. Patients were assigned to only one of the conditions of interest following priority order: CVST, VTE, and fracture. Fracture was selected because it was one of the most common conditions associated with the use of anticoagulants. Any patient who did not have CVST, VTE, or fracture was assigned to the “other” category. Within this “other” category, patients using anticoagulants who did not have a diagnostic code for CVST, VTE, or fracture were further explored for baseline conditions of pain, injury, arthropathy, and “other.” Among these latter conditions, patients could be counted in more than one condition category. The pediatric population after a Fontan procedure is of special interest. Given the current ongoing clinical development programs for anticoagulant therapies in the pediatric population, we also evaluated a cohort of patients who underwent the Fontan procedure regardless of anticoagulant use.
Outcome measures included VTE and major bleeding events without trauma based on implementing the algorithm by Cunningham et al
22
in claims data. Unadjusted incidence rates for VTE and major bleeding were generated for all observation days post-index treatment. Intention-to-treat incidence rates using all time at risk are presented in this report.Analyses
Counts were provided for categorical variables, and drug exposures were reported as means of duration of exposure, with a 30-day gap allowed for continuity determination. The incidence proportions were calculated as the number of events divided by the total number of patients at risk (expressed as per 1000 persons) and incidence rates as the number of events divided by the total follow-up time at risk (expressed as per 1000 person-years of exposure). Intention-to-treat analyses were conducted and defined as 1 day after the index date to the end of observation. Cases with events before the drug exposure of interest were included in analyses, as the events of interest could be an indication for treatment or secondary prevention but were not included in the relevant incidence rate calculation. Given the descriptive nature of the data and lack of adjustment between drug exposure populations, no statistical comparisons were made, and there was no intent to determine causality.
Results
Anticoagulant Use in the Pediatric Population
The number of eligible patients (aged <18 years) during the study period (2010–2019) was 28.4 million for CCAE, 12.1 million for MDCD, and 11.2 million for Optum. Overall use of anticoagulants was infrequent in the pediatric population. More than 49,000 patients in the pediatric population received anticoagulant therapies across the 3 databases (rivaroxaban, n = 778; apixaban, n = 216; all DOACs, n = 1442; warfarin, n = 1811; heparins, n = 46,229), in which the therapies were mainly dispensed to patients who were 12 to <18 years of age. The use of anticoagulants stratified according to age group and sex for each database is shown in Table I. All DOACs, and specifically rivaroxaban and apixaban, were used least often and primarily in patients aged 12 to <18 years. Across databases, among patients treated with warfarin, 15% to 23% were aged 2 to <7 years, 9% to 19% were aged 7 to <12 years, and the majority (>54%) were aged 12 to <18 years. Heparins were used most often overall and in the youngest patients. Anticoagulant use was generally balanced between male and female patients.
Table IOverall use of anticoagulants according to age group, sex, and database. Values are given as n (%) or mean (SD).
| Drug Exposure | CCAE | MDCD | Optum | |||
|---|---|---|---|---|---|---|
| Population | Drug Exposure Cumulative Days | Population | Drug Exposure Cumulative Days | Population | Drug Exposure Cumulative Days | |
| Rivaroxaban | 401 (100.0%) | 20,290 (50.6) | 217 (100.0%) | 16,962 (78.2) | 160 (100.0%) | 10,442 (65.3) |
| 1 to <2 y | – | – | 1 (0.5%) | 21 (21.0) | – | – |
| 2 to <7 y | 1 (0.2%) | 30 (30.0) | 6 (2.8%) | 412 (68.7) | – | – |
| 7 to <12 y | 8 (2.0%) | 429 (53.6) | 7 (3.2%) | 724 (103.4) | 3 (1.9%) | 297 (99.0) |
| 12 to <18 y | 392 (97.8%) | 19,831 (50.6) | 203 (93.5%) | 15,805 (77.9) | 157 (98.1%) | 10,145 (64.6) |
| Male | 184 (45.9%) | 8966 (48.7) | 126 (58.1%) | 8606 (68.3) | 77 (48.1%) | 4150 (53.9) |
| Female | 217 (54.1%) | 11,324 (52.2) | 91 (41.9%) | 8356 (91.8) | 83 (51.9%) | 6292 (75.8) |
| Apixaban | 121 (100.0%) | 9451 (78.1) | 60 (100.0%) | 3472 (57.9) | 35 (100.0%) | 2234 (63.8) |
| 1 to <2 y | – | – | 1 (1.7%) | 0 (0.0) | – | – |
| 2 to <7 y | 1 (0.8%) | 65 (65.0) | 1 (1.7%) | 30 (30.0) | – | – |
| 7 to <12 y | 3 (2.5%) | 118 (39.3) | 2 (3.3%) | 51 (25.5) | – | – |
| 12 to <18 y | 117 (96.7%) | 9268 (79.2) | 56 (93.3%) | 3391 (60.6) | 35 (100.0%) | 2234 (63.8) |
| Male | 62 (51.2%) | 3774 (60.9) | 35 (58.3%) | 1586 (45.3) | 22 (62.9%) | 1154 (52.5) |
| Female | 59 (48.8%) | 5677 (96.2) | 25 (41.7%) | 1886 (75.4) | 13 (37.1%) | 1080 (83.1) |
| All DOACs | 717 (100.0%) | 61,851 (86.3) | 423 (100.0%) | 47,645 (112.6) | 302 (100.0%) | 30,742 (101.8) |
| 1 to <2 y | – | – | 2 (0.5%) | 21 (10.5) | – | – |
| 2 to <7 y | 8 (1.1%) | 1054 (131.8) | 11 (2.6%) | 876 (79.6) | – | – |
| 7 to <12 y | 15 (2.1%) | 1322 (88.1) | 16 (3.8%) | 2538 (158.6) | 7 (2.3%) | 800 (114.3) |
| 12 to <18 y | 694 (96.8%) | 59,475 (85.7) | 394 (93.1%) | 44,210 (112.2) | 295 (97.7%) | 29,942 (101.5) |
| Male | 346 (48.3%) | 27,927 (80.7) | 233 (55.1%) | 23,512 (100.9) | 155 (51.3%) | 13,523 (87.2) |
| Female | 371 (51.7%) | 33,924 (91.4) | 190 (44.9%) | 24,133 (127.0) | 147 (48.7%) | 17,219 (117.1) |
| Warfarin | 903 (100.0%) | 188,267 (208.5) | 625 (100.0%) | 124,597 (199.4) | 283 (100.0%) | 63,140 (223.1) |
| <1 year | 2 (0.2%) | 432 (216.0) | 5 (0.8%) | 462 (92.4) | – | – |
| 1 to <2 y | 14 (1.6%) | 3841 (274.4) | 22 (3.5%) | 8126 (369.4) | 3 (1.1%) | 920 (306.7) |
| 2 to <7 y | 131 (14.5%) | 32,525 (248.3) | 142 (22.7%) | 25,231 (177.7) | 55 (19.4%) | 16,359 (297.4) |
| 7 to <12 y | 83 (9.2%) | 20,417 (246.0) | 118 (18.9%) | 35,547 (301.2) | 32 (11.3%) | 6072 (189.8) |
| 12 to <18 y | 673 (74.5%) | 131,052 (194.7) | 338 (54.1%) | 55,231 (163.4) | 193 (68.2%) | 39,789 (206.2) |
| Male | 491 (54.4%) | 107,333 (218.6) | 348 (55.7%) | 66,111 (190.0) | 163 (57.6%) | 39,261 (240.9) |
| Female | 412 (45.6%) | 80,934 (196.4) | 277 (44.3%) | 58,486 (211.1) | 120 (42.4%) | 23,879 (199.0) |
| Heparins | 20,593 (100.0%) | 390,894 (19.0) | 15,951 (100.0%) | 258,569 (16.2) | 9685 (100.0%) | 223,614 (23.1) |
| <1 year | 224 (1.1%) | 1728 (7.7) | 249 (1.6%) | 4978 (20.0) | 120 (1.2%) | 1657 (13.8) |
| 1 to <2 y | 1192 (5.8%) | 22,823 (19.1) | 1418 (8.9%) | 21,570 (15.2) | 560 (5.8%) | 9142 (16.3) |
| 2 to <7 y | 5223 (25.4%) | 141,426 (27.1) | 5267 (33.0%) | 101,663 (19.3) | 2605 (26.9%) | 87,701 (33.7) |
| 7 to <12 y | 4511 (21.9%) | 99,545 (22.1) | 3450 (21.6%) | 60,607 (17.6) | 2134 (22.0%) | 57,179 (26.8) |
| 12 to <18 y | 9443 (45.9%) | 125,372 (13.3) | 5567 (34.9%) | 69,751 (12.5) | 4266 (44.0%) | 67,935 (15.9) |
| Male | 11,085 (53.8%) | 221,180 (20.0) | 8571 (53.7%) | 151,031 (17.6) | 5212 (53.8%) | 128,269 (24.6) |
| Female | 9508 (46.2%) | 169,714 (17.8) | 7380 (46.3%) | 107,538 (14.6) | 4473 (46.2%) | 95,345 (21.3) |
DOACs = direct oral anticoagulants (rivaroxaban, apixaban, edoxaban, and dabigatran).
Databases include IBM MarketScan Commercial Claims and Encounters Database (CCAE), IBM MarketScan Multi-State Medicaid Database (MDCD), and Optum's de-identified Clinformatics Data Mart Database–Socio-Economic Status (Optum). A patient can be counted in more than one drug exposure category. Only continuous exposure is included; gaps >30 days are not included in the cumulative duration. Heparins include unfractionated heparin, enoxaparin, nadroparin, and dalteparin.
The use of anticoagulant therapy in the pediatric population with CVST, VTE, fracture, and all other conditions is shown in Table II. Overall, heparins were the most-used therapy in this study. Across databases, among the anticoagulants evaluated, heparins were used in 93% to 94% of pediatric patients, whereas all DOACs were used in 2% to 3%, and warfarin was used in 3% to 4% of pediatric patients. Of all DOAC use, VTE was the most common prior condition observed in 42% to 50% of pediatric patients, followed by other conditions, which represented 34% to 37% of all DOAC use. CVST was the least common condition observed, accounting for 2% to 3% of pediatric patients prescribed any DOACs. Of those prescribed warfarin, other conditions were most observed in 52% to 64% of patients, followed by VTE, which occurred in 30% to 40% of patients. CVST represented only 1% to 2% of all pediatric patients prescribed warfarin. Heparins were prescribed primarily for other conditions, observed in 93% to 94% of pediatric patients, whereas fractures represented 4% to 5%, VTE represented 2% to 3%, and CVST represented <1% of heparin use.
Table IIAnticoagulant use in pediatric populations with cerebral vein and sinus thrombosis (CVST), venous thromboembolism (VTE), or fracture during the study period (2010–2019). Values are given as n (%) or mean (SD).
| Drug Exposure | Type | CCAE | MDCD | Optum | ||||
|---|---|---|---|---|---|---|---|---|
| Population | Drug Exposure Cumulative Days | Population | Drug Exposure Cumulative Days | Population | Drug Exposure Cumulative Days | Cumulative Population | ||
| Rivaroxaban | CVST | 1 (0.2%) | 102 (102.0) | 2 (0.9%) | 724 (362.0) | 1 (0.6%) | 554 (554.0) | 4 (0.5%) |
| VTE | 120 (29.9%) | 12,371 (103.1) | 76 (35.0%) | 9715 (127.8) | 54 (33.8%) | 6668 (123.5) | 250 (32.1%) | |
| Fracture | 97 (24.2%) | 2206 (22.7) | 44 (20.3%) | 968 (22.0) | 37 (23.1%) | 1033 (27.9) | 178 (22.9%) | |
| Other | 183 (45.6%) | 5611 (30.7) | 95 (43.8%) | 5555 (58.5) | 68 (42.5%) | 2187 (32.2) | 346 (44.5%) | |
| Subtotal | 401 (100.0%) | 20,290 (50.6) | 217 (100.0%) | 16,962 (78.2) | 160 (100.0%) | 10,442 (65.3) | 778 (100.0%) | |
| Apixaban | CVST | 1 (0.8%) | 328 (328.0) | 0 (0.0%) | 0 (0.0) | 0 (0.0%) | 0 (0.0) | 1 (0.5%) |
| VTE | 47 (38.8%) | 6147 (130.8) | 26 (43.3%) | 2297 (88.3) | 15 (42.9%) | 1620 (108.0) | 88 (40.7%) | |
| Fracture | 27 (22.3%) | 922 (34.1) | 8 (13.3%) | 195 (24.4) | 2 (5.7%) | 51 (25.5) | 37 (17.1%) | |
| Other | 46 (38.0%) | 2054 (44.7) | 26 (43.3%) | 980 (37.7) | 18 (51.4%) | 563 (31.3) | 90 (41.7%) | |
| Subtotal | 121 (100.0%) | 9451 (78.1) | 60 (100.0%) | 3472 (57.9) | 35 (100.0%) | 2234 (63.8) | 216 (100.0%) | |
| All DOACs | CVST | 14 (2.0%) | 2963 (211.6) | 11 (2.6%) | 3210 (291.8) | 10 (3.3%) | 3587 (358.7) | 35 (2.4%) |
| VTE | 302 (42.1%) | 43,641 (144.5) | 212 (50.1%) | 32,104 (151.4) | 139 (46.0%) | 21,216 (152.6) | 653 (45.3%) | |
| Fracture | 134 (18.7%) | 3496 (26.1) | 57 (13.5%) | 1341 (23.5) | 46 (15.2%) | 1318 (28.7) | 237 (16.4%) | |
| Other | 267 (37.2%) | 11,751 (44.0) | 143 (33.8%) | 10,990 (76.9) | 107 (35.4%) | 4621 (43.2) | 517 (35.9%) | |
| Subtotal | 717 (100.0%) | 61,851 (86.3) | 423 (100.0%) | 47,645 (112.6) | 302 (100.0%) | 30,742 (101.8) | 1442 (100.0%) | |
| Warfarin | CVST | 14 (1.6%) | 3172 (226.6) | 7 (1.1%) | 983 (140.4) | 5 (1.8%) | 755 (151.0) | 26 (1.4%) |
| VTE | 350 (38.8%) | 62,395 (178.3) | 189 (30.2%) | 36,379 (192.5) | 114 (40.3%) | 21,596 (189.4) | 653 (36.1%) | |
| Fracture | 41 (4.5%) | 3431 (83.7) | 29 (4.6%) | 3752 (129.4) | 16 (5.7%) | 2190 (136.9) | 86 (4.7%) | |
| Other | 498 (55.1%) | 119,269 (239.5) | 400 (64.0%) | 83,483 (208.7) | 148 (52.3%) | 38,599 (260.8) | 1046 (57.8%) | |
| Subtotal | 903 (100.0%) | 188,267 (208.5) | 625 (100.0%) | 124,597 (199.4) | 283 (100.0%) | 63,140 (223.1) | 1811 (100.0%) | |
| Heparins | CVST | 37 (0.2%) | 693 (18.7) | 19 (0.1%) | 157 (8.3) | 17 (0.2%) | 429 (25.2) | 73 (0.2%) |
| VTE | 390 (1.9%) | 9748 (25.0) | 327 (2.1%) | 9588 (29.3) | 241 (2.5%) | 7026 (29.2) | 958 (2.1%) | |
| Fracture | 926 (4.5%) | 15,040 (16.2) | 569 (3.6%) | 3442 (6.0) | 467 (4.8%) | 8039 (17.2) | 1962 (4.2%) | |
| Other | 19,240 (93.4%) | 365,413 (19.0) | 15,036 (94.3%) | 245,382 (16.3) | 8960 (92.5%) | 208,120 (23.2) | 43,236 (93.5%) | |
| Subtotal | 20,593 (100.0%) | 390,894 (19.0) | 15,951 (100.0%) | 258,569 (16.2) | 9685 (100.0%) | 223,614 (23.1) | 46,229 (100.0%) | |
DOACs = direct oral anticoagulants (rivaroxaban, apixaban, edoxaban, and dabigatran).
Databases include IBM MarketScan Commercial Claims and Encounters Database (CCAE), IBM MarketScan Multi-State Medicaid Database (MDCD), and Optum's de-identified Clinformatics Data Mart Database–Socio-Economic Status (Optum). A patient can be counted in more than one drug exposure category. Only continuous exposure is included; gaps >30 days are not included in the cumulative duration. Conditions are identified in the time period 183 days prior, up to, and including the drug exposure start date. Heparins include unfractionated heparin, enoxaparin, nadroparin, and dalteparin. A patient can only be assigned to one of the conditions of interest by priority order: CVST, VTE, fracture, and other. Anyone who is neither CVST nor VTE nor fracture was assigned to “other,” which is described in additional detail in Table III.
The use of anticoagulants for other conditions (pain, injury, and arthropathy) is shown in Table III. These conditions are not mutually exclusive and indicate that pain (which was probably associated with the other conditions such as injury and arthropathy) was involved in most patients across all therapy types. Arthropathy and injury in pediatric patients generally accounted for a large proportion of DOAC use (>50%) and of warfarin and heparin use (up to 46%).
Table IIIPain, injury, arthropathy, and other conditions among patients in the “other” category during the study period (2010–2019). Data are n (%).
| Drug Exposure | Condition | CCAE | MDCD | Optum |
|---|---|---|---|---|
| Population | Population | Population | ||
| Rivaroxaban | Pain | 144 (81.8%) | 55 (60.4%) | 55 (87.3%) |
| Injury | 139 (79.0%) | 53 (58.2%) | 52 (82.5%) | |
| Arthropathy | 138 (78.4%) | 45 (49.5%) | 52 (82.5%) | |
| Other | 18 (10.2%) | 28 (30.8%) | 6 (9.5%) | |
| Apixaban | Pain | 37 (80.4%) | 18 (69.2%) | 12 (66.7%) |
| Injury | 31 (67.4%) | 14 (53.8%) | 11 (61.1%) | |
| Arthropathy | 26 (56.5%) | 14 (53.8%) | 10 (55.6%) | |
| Other | 8 (17.4%) | 6 (23.1%) | 3 (16.7%) | |
| All DOACs | Pain | 209 (78.3%) | 88 (61.5%) | 82 (76.6%) |
| Injury | 189 (70.8%) | 74 (51.7%) | 75 (70.1%) | |
| Arthropathy | 180 (67.4%) | 62 (43.4%) | 69 (64.5%) | |
| Other | 41 (15.4%) | 43 (30.1%) | 18 (16.8%) | |
| Warfarin | Pain | 230 (46.2%) | 153 (38.2%) | 74 (50.0%) |
| Injury | 158 (31.7%) | 183 (45.8%) | 47 (31.8%) | |
| Arthropathy | 63 (12.7%) | 32 (8.0%) | 21 (14.2%) | |
| Other | 212 (46.2%) | 138 (34.5%) | 61 (41.2%) | |
| Heparins | Pain | 9556 (52.0%) | 7065 (48.1%) | 4474 (53.9%) |
| Injury | 5095 (27.7%) | 4115 (28.0%) | 2281 (27.5%) | |
| Arthropathy | 2739 (14.9%) | 1558 (10.6%) | 1212 (14.6%) | |
| Other | 7373 (40.1%) | 6131 (41.8%) | 3233 (39.0%) |
DOACs = direct oral anticoagulants (rivaroxaban, apixaban, edoxaban, and dabigatran).
Databases include IBM MarketScan Commercial Claims and Encounters Database (CCAE), IBM MarketScan Multi-State Medicaid Database (MDCD), and Optum's de-identified Clinformatics Data Mart Database–Socio-Economic Status (Optum). A patient can be counted in more than one condition category (pain, injury, and arthropathy). Patients in the “other” category do not have pain, injury, or arthropathy. Conditions are identified in the time period 183 days prior, up to and including the drug exposure start date. The denominator for the proportions displayed is the total from the “other” category in Table II.
Unadjusted Major Bleeding and VTE Incidence Rates
The unadjusted incidences of major bleeding and VTE in patients receiving anticoagulant therapies stratified according to anticoagulant, using all time at risk during the study period, are shown in Table IV. Rates of major bleeding per 1000 person-years of exposure ranged from 0 to 3.0 for rivaroxaban, 0 to 5.1 for apixaban, and 0 to 4.9 for all DOACs. Rates of major bleeding per 1000 person-years of exposure ranged from 4.3 to 6.7 for warfarin and 3.7 to 4.6 for heparins. Rates of VTE per 1000 person-years of exposure ranged from 15.0 to 24.9 for rivaroxaban, 22.9 to 48.5 for apixaban, and 30.8 to 34.0 for all DOACs. Rates of VTE per 1000 person-years of exposure ranged from 21.6 to 46.2 for warfarin and 6.0 to 7.3 for heparins. Supplemental Tables V and VI (see the online version at doi:10.1016/j.clinthera.2021.09.021) provide unadjusted incidence rates for each of the 5-year periods of the study, 2010 to 2014 and 2015 to 2019, respectively.
Table IVIntention-to-treat incidence rates of major bleeding (MB) without trauma and venous thromboembolism (VTE) using all time at risk during the study period (2010–2019) in the pediatric population.
| Database | Drug Exposure | Outcome | Patients (N) | Events (N) | Person-Years of Exposure | Proportion per 1000 Persons | Rate per 1000 Person-Years of Exposure |
|---|---|---|---|---|---|---|---|
| CCAE | Rivaroxaban | MB | 400 | 1 | 856 | 2.5 | 1.2 |
| VTE | 400 | 15 | 825 | 37.5 | 18.2 | ||
| Apixaban | MB | 121 | 1 | 197 | 8.3 | 5.1 | |
| VTE | 121 | 9 | 185 | 74.4 | 48.5 | ||
| All DOACs | MB | 716 | 2 | 1440 | 2.8 | 1.4 | |
| VTE | 716 | 42 | 1362 | 58.7 | 30.8 | ||
| Warfarin | MB | 903 | 12 | 2469 | 13.3 | 4.9 | |
| VTE | 903 | 65 | 2283 | 72 | 28.5 | ||
| Heparins | MB | 20,557 | 164 | 44,203 | 8 | 3.7 | |
| VTE | 20,557 | 288 | 43,944 | 14 | 6.6 | ||
| MDCD | Rivaroxaban | MB | 217 | 0 | 403 | 0 | 0 |
| VTE | 217 | 6 | 399 | 27.6 | 15 | ||
| Apixaban | MB | 60 | 0 | 71 | 0 | 0 | |
| VTE | 60 | 3 | 68 | 50 | 43.8 | ||
| All DOACs | MB | 423 | 0 | 756 | 0 | 0 | |
| VTE | 423 | 23 | 727 | 54.4 | 31.6 | ||
| Warfarin | MB | 625 | 8 | 1881 | 12.8 | 4.3 | |
| VTE | 625 | 39 | 1804 | 62.4 | 21.6 | ||
| Heparins | MB | 15,934 | 181 | 44,914 | 11.4 | 4 | |
| VTE | 15,934 | 323 | 44,523 | 20.3 | 7.3 | ||
| Optum | Rivaroxaban | MB | 159 | 1 | 331 | 6.3 | 3 |
| VTE | 159 | 8 | 321 | 50.3 | 24.9 | ||
| Apixaban | MB | 35 | 0 | 45 | 0 | 0 | |
| VTE | 35 | 1 | 43 | 28.6 | 22.9 | ||
| All DOACs | MB | 301 | 3 | 614 | 10 | 4.9 | |
| VTE | 301 | 20 | 588 | 66.4 | 34 | ||
| Warfarin | MB | 282 | 5 | 747 | 17.7 | 6.7 | |
| VTE | 282 | 31 | 671 | 109.9 | 46.2 | ||
| Heparins | MB | 9665 | 103 | 22,385 | 10.7 | 4.6 | |
| VTE | 9665 | 134 | 22,270 | 13.9 | 6 |
DOACs = direct oral anticoagulants (rivaroxaban, apixaban, edoxaban, and dabigatran).
Databases include IBM MarketScan Commercial Claims and Encounters Database (CCAE), IBM MarketScan Multi-State Medicaid Database (MDCD), and Optum's de-identified Clinformatics Data Mart Database–Socio-Economic Status (Optum).
Fontan Procedure
Record of a Fontan procedure was found in 1971, 1785, and 838 pediatric patients in the CCAE, MDCD, and Optum databases, respectively, during the study period. All patients were identified regardless of anticoagulant use. Across all databases and among the Fontan procedure subpopulation, heparins were prescribed in 33% of patients (n = 1500 [CCAE, n = 558; MDCD, n = 650; Optum, n = 292]), followed by warfarin in 3% of patients (n = 156 [CCAE, n = 72; MDCD, n = 49; Optum, n = 35]). Among the Fontan subpopulation, the mean number of days of follow-up time after the procedure was 941 in CCAE, 1059 in MDCD, and 920 in Optum. Across all Fontan patients, the incidence rates of VTE ranged from 4.3 to 8.7, and rates of major bleeding ranged from 2.7 to 3.4 per 1000 person-years of exposure (Table V).
Table VIntention-to-treat incidence rates of major bleeding (MB) without trauma and venous thromboembolism (VTE) during the study period (2010–2019) in the post–Fontan procedure population.
| Database | Drug Exposure | Outcome | Patients (N) | Events (N) | Person-Years of Exposure | Proportion per 1000 Persons | Rate per 1000 Person-Years of Exposure |
|---|---|---|---|---|---|---|---|
| CCAE | All DOACs | MB | 3 | 0 | 7 | 0 | 0 |
| VTE | 3 | 0 | 7 | 0 | 0 | ||
| Warfarin | MB | 72 | 0 | 218 | 0 | 0 | |
| VTE | 72 | 3 | 207 | 41.7 | 14.5 | ||
| Heparins | MB | 558 | 1 | 1241 | 1.8 | 0.8 | |
| VTE | 558 | 3 | 1231 | 5.4 | 2.4 | ||
| All anticoagulants | MB | 1971 | 17 | 5029 | 8.6 | 3.4 | |
| VTE | 1971 | 43 | 4963 | 21.8 | 8.7 | ||
| MDCD | All DOACs | MB | 0 | 0 | 0 | NA | NA |
| VTE | 0 | 0 | 0 | NA | NA | ||
| Warfarin | MB | 49 | 1 | 137 | 20.4 | 7.3 | |
| VTE | 49 | 4 | 135 | 81.6 | 29.5 | ||
| Heparins | MB | 650 | 4 | 1728 | 6.2 | 2.3 | |
| VTE | 650 | 6 | 1719 | 9.2 | 3.5 | ||
| All anticoagulants | MB | 1785 | 14 | 5154 | 7.8 | 2.7 | |
| VTE | 1785 | 24 | 5129 | 13.4 | 4.7 | ||
| Optum | All DOACs | MB | 1 | 0 | 3 | 0 | 0 |
| VTE | 1 | 0 | 3 | 0 | 0 | ||
| Warfarin | MB | 35 | 1 | 79 | 28.6 | 12.6 | |
| VTE | 35 | 1 | 77 | 28.6 | 12.9 | ||
| Heparins | MB | 292 | 0 | 716 | 0 | 0 | |
| VTE | 292 | 0 | 716 | 0 | 0 | ||
| All anticoagulants | MB | 838 | 7 | 2106 | 8.4 | 3.3 | |
| VTE | 838 | 9 | 2100 | 10.7 | 4.3 |
DOACs = direct oral anticoagulants (rivaroxaban, apixaban, edoxaban, and dabigatran); NA = not applicable.
No patients had events with rivaroxaban or apixaban. Databases include IBM MarketScan Commercial Claims and Encounters Database (CCAE), IBM MarketScan Multi-State Medicaid Database (MDCD), and Optum's de-identified Clinformatics Data Mart Database–Socio-Economic Status (Optum).
Discussion
This analysis characterizes the use of anticoagulants in pediatric patients using large US health care claims databases and illustrates the potential to explore clinical outcomes of interest related to index conditions, including those that require anticoagulant therapies. Although the claims data do not provide the reasons why these medications were prescribed, the diagnostic codes associated with clinical encounters allow an inference as to why they might have been prescribed. DOACs are not yet approved by the US Food and Drug Administration in pediatrics, and we found that use of this class, in general, was uncommon in the pediatric population, particularly among children younger than 12 years of age. Heparins were the most commonly used anticoagulants, followed by warfarin. Among patients receiving DOACs, VTE was the most common condition reported in the 6 months before anticoagulant use.
The present analysis determined the incidence rates of VTE and major bleeding for each type of anticoagulant. These data are not intended to imply causality or to be used comparatively, given the lack of adjustments for confounding variables between drug exposure populations and the inability to differentiate between new and recurrent events. However, these real-world data serve as a useful benchmark of current clinical practices for children receiving anticoagulation therapy, and they thereby increase our understanding of the use of anticoagulation therapy in children with at-risk conditions.
Our analysis identified 4594 children with a history of a Fontan procedure. Anticoagulation therapy was used in a similar manner in these patients compared with the overall population. Few patients who had the Fontan procedure received DOACs. Heparins were most often prescribed. The CCAE, MDCD, and Optum databases used in this study do not capture information about ASA because this medication is not obtained through insurance in the United States. Meta-analyses in the literature of patients after a Fontan procedure have found reduced thromboembolic events with prophylactic therapy (ASA or warfarin) versus no prophylactic therapy.
23
No difference between anticoagulation therapy and antiplatelet therapy has been observed in this population.11
,24
Most anticoagulants currently used in clinical practice are not approved by the US Food and Drug Administration in the pediatric population, and their use in children is based on adult data and the limited available pediatric data. Therefore, in our analysis, the condition before the index anticoagulant exposure may be associated with the use of the anticoagulant, but such correlations should be carefully interpreted. In addition to the conditions of CVST, VTE, and fractures, anticoagulant use was identified with other common conditions, including pain, injury, and arthropathy. Although the indication for prescribing these medications cannot be determined from the CCAE, MDCD, and Optum databases, it may be presumed that anticoagulants were prescribed for thrombosis treatment or, more likely, for thromboprophylaxis. An example would be thromboprophylaxis related to surgery or injury that led to immobility. As the publication of clinical trial results for DOACs in the pediatric population are expected in the next 2 years, the present real-world analysis represents an important view of current pediatric clinical practice and may be helpful in evaluating future changes in utilization and outcome events across anticoagulants.
This study has strengths with respect to the use of 3 large US health claims databases, as they provide a less homogeneous population than would be observed in a necessarily small pediatric clinical trial. The databases used in this study include records of diagnoses and prescriptions written and provide rich and broad data that are less expensive and more feasible to obtain compared with a patient registry study. Health claims databases allow for visibility into changing trends in prescription practices and outcomes.
The present study is limited by factors inherently associated with health claims databases. Data are only captured when a patient seeks care, and individuals who lack or have insufficient medical insurance could be underrepresented in the data. Challenges in interpreting the available data arise from the inability of the databases to capture the reasons why medications were prescribed, whether the medication was used therapeutically or for prophylaxis, and whether events were new or recurrent. In addition, information about medications obtained without insurance (eg, ASA) is not available, and in-hospital drug exposures are incomplete, with drug-dispensing information limited to the outpatient setting only. Claims are adjudicated by payors, and a time lag exists from the actual occurrence; coding can be distorted by the requirement to code for reimbursement. Claims data do not provide a complete prevalence of patients. These numbers describe the populations captured by these respective databases, and care should be taken when generalizing findings to the broader US population. The unadjusted incidence proportions and rates for the individual drugs are not intended for statistical comparisons or to imply causation. This study also investigated a subpopulation of pediatric anticoagulant users who had a prior Fontan procedure. This procedure is typically provided to young patients (<4 years old), but the anticoagulant exposures evaluated were predominantly identified in patients aged 12 to 18 years, and thus caution should be exercised when interpreting the results in this subpopulation as the drug exposures may not be temporally associated with the procedure itself.
Conclusions
This real-world analysis of US claims data found that pediatric patients currently have limited exposure to anticoagulant therapy, particularly DOACs. Heparins are used most often in this population and particularly in those aged <12 years. Our study investigated the following conditions before anticoagulation therapy in pediatric patients: CVST, VTE, fracture, pain, injury, and arthropathy. Among patients prescribed an anticoagulant, a subpopulation of those who had a Fontan procedure were evaluated for VTE and major bleeding events. Incidence of VTE was higher compared with the incidence of major bleeding.
In the next 2 years, approvals of pediatric product labeling updates are anticipated for DOACs, as randomized clinical trials are completed in the pediatric population. Because randomized controlled trials in children are necessarily limited in sample size, real-world data such as those from the present analyses may serve as a useful source to generate real-world evidence to help inform the safe and effective use of these therapies. In this regard, they are complementary to the results from randomized clinical trials. Continued use of real-world data analyses such as in the present study may also provide important insights into clinical practice trends, challenges, and opportunities that serve to improve the use of anticoagulant therapies in the pediatric population in the United States.
Data Statement
De-identified data were available to the authors via license between IBM MarketScan and Optum, a commercial data provider in the United States, and Janssen Pharmaceuticals. As such, the authors cannot provide the raw data; however, interested researchers may access the data through IBM and Optum, and the methods applied in the article would allow study replication.
Disclosures
This study was funded by Janssen Research & Development, LLC. J.H., L.M.P., C.P., H.T., C.K., and Z.Y. are full-time employees of Janssen Research & Development, LLC, and P.S. is a full-time employee of Johnson & Johnson; therefore, all of these authors may be stockholders of Johnson & Johnson. A.D.M. has served on scientific advisory boards for AstraZeneca and Janssen and has served as a consultant to Medtronic and Stasys. B.W.M. has served on scientific advisory boards for Janssen, Esperion, and Chiesi; has served as data safety monitoring board chair for Amryt Pharma; and has served as an investigator for Mezzion and Janssen. The study sponsor was involved in the study design; in the collection, analysis, and interpretation of the data; in the writing of the manuscript; and in the decision to submit the manuscript for publication.
Acknowledgments
The authors wish to thank Hernando Patino, of Johnson & Johnson, for his contributions to the study design. Medical writing support was provided by Michelle McDermott, PharmD, of Cello Health Communications/MedErgy, and was funded by Janssen Scientific Affairs, LLC.
Appendix. Supplementary materials
References
- American Society of Hematology 2018 guidelines for management of venous thromboembolism: treatment of pediatric venous thromboembolism.Blood Adv. 2018; 2: 3292-3316
- Safety of dabigatran etexilate for the secondary prevention of venous thromboembolism in children.Blood. 2020; 135: 491-504
- Management of thrombosis in children and neonates: practical use of anticoagulants in children.Hematology Am Soc Hematol Educ Program. 2018; 2018: 399-404
- Neonatal venous thromboembolism.Front Pediatr. 2017; 5: 136
- Rivaroxaban, a direct factor Xa inhibitor, versus acetylsalicylic acid as thromboprophylaxis in children post-Fontan procedure: rationale and design of a prospective.Am Heart J. 2019; 213 (randomized trial (the UNIVERSE study)): 97-104
- Thromboembolism and anticoagulation after Fontan surgery.Ann Pediatr Cardiol. 2016; 9: 236-240
- Intracardiac thrombus formation after the Fontan operation.J Thorac Cardiovasc Surg. 2000; 119: 745-752
- Transesophageal echocardiography detects thrombus formation not identified by transthoracic echocardiography after the Fontan operation.J Am Coll Cardiol. 1991; 18: 1733-1737
- Transesophageal echocardiography in evaluation and management after a Fontan procedure.J Am Coll Cardiol. 1991; 17: 1152-1160
- Factors associated with thrombotic complications after the Fontan procedure: a secondary analysis of a multicenter, randomized trial of primary thromboprophylaxis for 2 years after the Fontan procedure.J Am Coll Cardiol. 2013; 61: 346-353
- A multicenter, randomized trial comparing heparin/warfarin and acetylsalicylic acid as primary thromboprophylaxis for 2 years after the Fontan procedure in children.J Am Coll Cardiol. 2011; 58: 645-651
- Prevention and treatment of thrombosis in pediatric and congenital heart disease: a scientific statement from the American Heart Association.Circulation. 2013; 128: 2622-2703
- Executive summary: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines.Chest. 2012; 141: 7S-47S
- Antithrombotic therapy in neonates and children: antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines.Chest. 2012; 141: e737S-e801S
- Treatment of venous thromboembolism in pediatric patients.Blood. 2020; 135: 335-343
- Venous thromboembolism in children: from diagnosis to management.Int J Environ Res Public Health. 2020; 17: 4993
- Low antithrombin levels in neonates and infants undergoing congenital heart surgery result in more red blood cell and plasma transfusion on cardiopulmonary bypass.Transfusion. 2020; 60: 2841-2848
- Challenges in maintaining the hemostatic balance in children undergoing extracorporeal membrane oxygenation: a systematic literature review.Front Pediatr. 2020; 8: 612467
- Rivaroxaban compared with standard anticoagulants for the treatment of acute venous thromboembolism in children: a randomised, controlled, phase 3 trial.Lancet Haematol. 2020; 7: e18-e27
- Documentation for the database transformations to the CDM.2020 (Available at:) (Accessed October 22, 2020)
- Common Data Model.2020 (Available at:) (Accessed October 22, 2020)
- An automated database case definition for serious bleeding related to oral anticoagulant use.Pharmacoepidemiol Drug Saf. 2011; 20: 560-566
- Strategies for thromboprophylaxis in Fontan circulation: a meta-analysis.Heart. 2015; 101: 1731-1737
- Antiplatelet versus anticoagulation therapy after extracardiac conduit Fontan: a systematic review and meta-analysis.Pediatr Cardiol. 2011; 32: 32-39
Article info
Publication history
Published online: December 03, 2021
Accepted:
September 25,
2021
Identification
Copyright
© 2021 Janssen Research & Development, LLC. Published by Elsevier Inc.
User license
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) | How you can reuse 
Elsevier's open access license policy
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)
Permitted
For non-commercial purposes:
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article (private use only, not for distribution)
- Reuse portions or extracts from the article in other works
Not Permitted
- Sell or re-use for commercial purposes
- Distribute translations or adaptations of the article
Elsevier's open access license policy
