Abstract
Background
The European rivaroxaban post-authorization safety study evaluated bleeding risk among patients initiated on rivaroxaban or vitamin K antagonists for the treatment and secondary prevention of venous thromboembolism in routine clinical practice.
Methods
Cohorts were created using electronic healthcare databases from the UK, the Netherlands, Germany and Sweden. Patients with a first prescription of rivaroxaban or vitamin K antagonist during the period from December 2011 (in the UK, January 2012) to December 2017 (in Germany, December 2016) for venous thromboembolism indication, with no record of atrial fibrillation or recent cancer history, were observed until the occurrence of each safety outcome (hospitalization for intracranial, gastrointestinal, urogenital or other bleeding), death or study end (December 2018; in Germany, December 2017). Crude incidence rates of each outcome per 100 person-years were computed.
Results
Overall, 44 737 rivaroxaban and 45 842 vitamin K antagonist patients were enrolled, mean age, 59.9–63.8 years. Incidence rates were similar between rivaroxaban and vitamin K antagonist users with some exceptions, including higher incidence rates for gastrointestinal bleeding in rivaroxaban users than in vitamin K antagonist users. Among rivaroxaban users, mortality and bleeding risk generally increased with age, renal impairment and diabetes.
Conclusions
This study provides further data from routine clinical practice that broadly support safety profile of rivaroxaban for VTE indication and complement findings from previous randomized clinical trials.
Introduction
Venous thromboembolism (VTE), clinically presenting as deep vein thrombosis (DVT) or pulmonary embolism (PE), is associated with significant morbidity and mortality. Epidemiological modelling for six European countries estimated nearly 762 000 symptomatic VTE events/year and over 370 000 VTE-related deaths/year [1]. Risk factors for VTE (other than major surgery and trauma) [2] include increasing age [3], cancer [3], hospitalization [4] and previous history of VTE [5].
Anticoagulants are the cornerstone of treatment and secondary prevention of VTE [3, 6]. The choice of anticoagulant is driven by patient characteristics, including comorbidities and comedications, and practice guidelines for the relevant healthcare system [3, 6]. The primary safety concern for all anticoagulants is the increased risk of bleeding, especially events requiring hospitalization. Understanding the likely bleeding risk for a patient is a key component of the benefit–risk evaluation of anticoagulant treatment when deciding on a drug and treatment duration.
Rivaroxaban is an oral, direct, factor Xa inhibitor that has been approved for several indications, including the treatment and secondary prevention of VTE in adults [7, 8]. In clinical trials in patients with VTE, rivaroxaban was associated with a lower rate of major bleeding than the standard of care (enoxaparin and either warfarin or acenocoumarol) [9]. Shortly after the first approval of rivaroxaban in Europe in 2011, a pharmacoepidemiological post-authorization safety study (PASS) programme was initiated to monitor patterns of rivaroxaban use, patient characteristics and safety and effectiveness outcomes in first-time users of rivaroxaban and vitamin K antagonists (VKAs) treated for the condition in routine clinical practice [10]. One element of the rivaroxaban PASS incorporates four observational cohort studies, collectively covering the UK, the Netherlands, Germany and Sweden over 5 years.
The objective of the present study was to investigate the safety outcomes (intracranial, gastrointestinal, urogenital and other bleeding events leading to hospitalization), selected based on the outcomes included in pivotal randomized controlled trials (RCTs) of rivaroxaban [11, 12]. Additionally, we evaluated mortality risk and bleeding outcomes according to old age, renal function and diabetes among rivaroxaban users.
Methods
Study design
The study design of this PASS has been previously published in detail [10]. Study cohorts were created using healthcare databases from four European countries. Enrolment of patients aged 2 years or older who were first-time users of rivaroxaban or VKA for an indication of the treatment and secondary prevention of VTE began on 9 December 2011 (Germany, Netherlands and Sweden) or 1 January 2012 (UK) and continued until 31 December 2016 (Germany) or 31 December 2017 (UK, Netherlands and Sweden); the observation period ended 1 year after the end of the enrolment period. Patients who previously received treatment with other forms of anticoagulant therapy were eligible for inclusion (S1 Table).
The start date for each patient was defined as the date of first prescription/dispensation of study drug (rivaroxaban or VKA) during the enrolment period. The comparator study drug was defined as the most widely used VKA(s) in each of the respective countries (UK and Sweden, warfarin; Germany, phenprocoumon; Netherlands, acenocoumarol or phenprocoumon). An indication of the treatment and secondary prevention of VTE was assigned when a diagnostic code for DVT and/or PE was recorded close to the prescription date (3 months prior), in the absence of an atrial fibrillation code at any time point.
VKA users were included in the PASS to provide contemporary context for rivaroxaban use because VKAs were the standard of care at the time of study initiation. The study was not designed for direct statistical comparisons of incidence rates (IRs) among anticoagulants or countries and therefore only absolute risks are presented. Patients with a recent history of cancer (i.e. a cancer code received from 3 years before to 1 month after the start date was used as a proxy for active cancer) were excluded from the present analyses because patients with active cancer are a heterogeneous group that warrants specific consideration owing to a variably increased risk of thromboembolic and bleeding events, which makes the choice of anticoagulant and duration of treatment particularly challenging [13–15].
Data sources
The data sources for the PASS included the following: the IQVIA Medical Research Data (IMRD), incorporating The Health Improvement Network(THIN), a Cegedim database; the German Pharmacoepidemiological Research Database (GePaRD); the PHARMO Database Network, Netherlands (for the general practitioner sub-cohort in this analysis, data were extracted from the Out-patient Pharmacy Database, Hospitalisation Database and General Practitioner Database); and Swedish nationwide health registries [10]. Approval of individual study protocols were granted by the appropriate research ethics committees and regulatory authorities (S2 Table). The four studies are registered separately on the ENCePP (European Network of Centres for Pharmacoepidemiology and Pharmacovigilance) electronic register (EUPAS11299, EUPAS11141, EUPAS11145, and EUPAS9895) and the ClinicalTrials.gov website (NCT01947998, NCT01947985, NCT01947959, and NCT02468102).
Outcomes
The definitions of outcomes were based on specific diagnostic procedure codes (Read Codes in the UK, or International Statistical Classification of Diseases and Related Health Problems [ICD] 9th Revision or ICD 10th Revision in Germany, the Netherlands and Sweden) and were harmonized between databases as much as possible for the purpose of this analysis; studies that detail the design and validation of this methodology have previously been published [10, 16]. The evaluated safety outcomes included bleeding events that led to hospitalization (as measure of severity), categorized as intracranial haemorrhage (intracerebral haemorrhage, subarachnoid haemorrhage or subdural haematoma), gastrointestinal bleeding (upper and lower gastrointestinal tract), urogenital bleeding and bleeding events at other sites [17]. All-cause mortality was also evaluated among rivaroxaban users.
Statistical methods
The data from each country were analysed separately and anonymously, authors have no access that could identify patients participants. Data were not pooled between countries due to databases differences and to permit transparency of findings. Independent follow-ups were performed for each outcome, and patients were censored at whichever came first of withdrawal from the database; first occurrence of the safety outcome of interest; treatment switch or discontinuation from the initial drug (with a 30-day grace period after the end of supply of the study drug); death or the end of the observation period. The safety outcomes for first-time users of rivaroxaban or VKAs during their first episode of treatment were analysed as unadjusted IRs with 95% confidence intervals (CIs) computed as the number of events per 100 person-years. IRs of bleeding and all-cause mortality were also calculated in specific high-risk subgroups of rivaroxaban users, including patients with impaired renal function, patients aged 75 years or older and patients with type 1 or type 2 diabetes. No comparative statistical analyses of unadjusted IRs of bleeding outcomes between rivaroxaban and VKAs were conducted because these drugs were likely prescribed to groups of patients with different characteristics that cannot be fully adjusted for in the analyses.
Results
Study population
The analyses included 44 737 patients treated with rivaroxaban and 45 842 treated with VKAs (mean age, 59.9–63.8 years; approximately 1:1 ratio of male to female patients). Baseline characteristics for patients were broadly similar across the countries (Table 1). The proportion of patients with a history of bleeding across cohorts ranged from 0.5% (intracranial bleeding in VKA-treated patients in Germany) to 11.4% (urogenital bleeding in rivaroxaban-treated patients in the UK). Most patients were naive to oral anticoagulant therapy at the start date (rivaroxaban users, 87.7–98.5%; VKA users, 99.0–99.6%).
Table 1. General characteristics of study patients: First-time users of rivaroxaban or VKA for the treatment and secondary prevention of VTE.
| UK | Netherlands | Germany | Sweden | |||||
|---|---|---|---|---|---|---|---|---|
| Rivaroxaban | VKA(s) | Rivaroxaban | VKA(s) | Rivaroxaban | VKA(s) | Rivaroxaban | VKA(s) | |
| N = 5680 | N = 4636 | N = 586 | N = 2617 | N = 25 914 | N = 20 502 | N = 12 557 | N = 18 087 | |
| Sex | ||||||||
| Male | 2809 (49.5) | 2335 (50.4) | 274 (46.8) | 1278 (48.8) | 11 860 (45.8) | 9290 (45.3) | 6319 (50.3) | 8913 (49.3) |
| Female | 2871 (50.5) | 2301 (49.6) | 312 (53.2) | 1339 (51.2) | 14 054 (54.2) | 11 212 (54.7) | 6238 (49.7) | 9174 (50.7) |
| Age at first prescription/dispensation, years, mean ± SD | 62.3 ± 16.5 | 62.1 ± 16.1 | 60.7 ± 16.2 | 59.9 ± 16.2 | 60.9 ± 17.4 | 62.4 ± 16.7 | 62.1 ± 17.2 | 63.8 ± 17.2 |
| Age group | ||||||||
| ≤ 70 years | 3468 (61.1) | 2931 (63.2) | 389 (66.4) | 1881 (71.9) | 16 147 (62.3) | 12 049 (58.8) | 7770 (61.9) | 10 511 (58.1) |
| > 70 years | 2212 (39.0) | 1705 (36.8) | 197 (33.6) | 736 (28.1) | 9767 (37.7) | 8453 (41.2) | 4787 (38.1) | 7576 (41.9) |
| Naive to OAC at first prescription/dispensationa | 4980 (87.7) | 4618 (99.6) | 545 (93.0) | 2606 (99.6) | 22 809 (88.0) | 20 323 (99.1) | 12 372 (98.5) | 17 911 (99.0) |
| Number of contacts with GP during the 12 months before the start date | ||||||||
| 0–9 | 1310 (23.1) | 1176 (25.4) | 496 (84.6) | 2252 (86.1) | –b | –b | 11 716 (93.3)c | 16 837 (93.1)c |
| 10–19 | 1924 (33.9) | 1715 (37.0) | 79 (13.5) | 317 (12.1) | –b | –b | 715 (5.7)c | 1033 (5.7)c |
| ≥ 20 | 2446 (43.1) | 1745 (37.6) | 11 (1.9) | 48 (1.8) | –b | –b | 126 (1.0)c | 217 (1.2)c |
| Number of hospitalizations during the 12 months before the start date | ||||||||
| None | 2692 (47.4) | 1901 (41.0) | 386 (65.9) | 1119 (42.8) | 6897 (26.6) | 4346 (21.2) | 5365 (42.7) | 5410 (29.9) |
| 1 | 1361 (24.0) | 1224 (26.4) | 146 (24.9) | 1002 (38.3) | 11 984 (46.2) | 9875 (48.2) | 4660 (37.1) | 7669 (42.4) |
| ≥ 2 | 1627 (28.6) | 1511 (32.6) | 54 (9.2) | 496 (19.0) | 7033 (27.1) | 6281 (30.6) | 2532 (20.2) | 5008 (27.7) |
| Lifestyle characteristics | ||||||||
| Smokers | 1124 (19.8) | 831 (17.9) | 37 (6.3) | 182 (7.0) | –d | –d | 244 (1.9) | 356 (2.0) |
| Obesity or BMI > 30 | 2211 (38.9) | 1818 (39.2) | 99 (16.9) | 341 (13.0) | 9194 (35.5) | 7473 (36.5) | 568 (4.5) | 855 (4.7) |
| Polypharmacye | ||||||||
| ≥ 5 medications | 3178 (56.0) | 2494 (53.8) | 239 (40.8) | 1421 (54.3) | 8395 (32.4) | 7317 (35.7) | –e | –e |
| Medications of interest prescribed up to 90 days before or on the start date | ||||||||
| Antiplatelets | 1068 (18.8) | 906 (19.5) | 54 (9.2) | 254 (9.7) | 1295 (5.0) | 1135 (5.5) | 1204 (9.6) | 2242 (12.4) |
| NSAIDs | 825 (14.5) | 721 (15.6) | 54 (9.2) | 192 (7.3) | 6756 (26.1) | 5147 (25.1) | 898 (7.2) | 1154 (6.4) |
| Antidiabetic agents | 529 (9.3) | 428 (9.2) | 38 (6.5) | 201 (7.7) | 1852 (7.1) | 1622 (7.9) | 771 (6.1) | 1315 (7.3) |
| Oral steroids | 682 (12.0) | 551 (11.9) | 42 (7.2) | 247 (9.4) | 2392 (9.2) | 2324 (11.3) | 1315 (10.5) | 2264 (12.5) |
| PPI | 2123 (37.4) | 1617 (34.9) | 160 (27.3) | 846 (32.2) | 8670 (33.5) | 6985 (34.1) | 2122 (16.9) | 3413 (18.9) |
| Medical history: past medical events and comorbidities occurring any time before the start date | ||||||||
| Intracranial bleeding (intracerebral, other) | 58 (1.0) | 53 (1.1) | 4 (0.7) | 21 (0.8) | 229 (0.9) | 104 (0.5) | 229 (1.8) | 296 (1.6) |
| Gastrointestinal bleeding | 638 (11.2) | 470 (10.1) | 8 (1.4) | 23 (0.9) | 447 (1.7) | 334 (1.6) | 370 (2.9) | 544 (3.0) |
| Urogenital bleeding | 649 (11.4) | 526 (11.3) | 7 (1.2) | 24 (0.9) | 218 (0.8) | 138 (0.7) | 662 (5.3) | 849 (4.7) |
| Ischaemic stroke | 276 (4.9) | 213 (4.6) | 9 (1.5) | 58 (2.2) | 1821 (7.0) | 1340 (6.5) | 507 (4.0) | 915 (5.1) |
| MI | 277 (4.9) | 258 (5.6) | 35 (6.0) | 115 (4.4) | 1599 (6.2) | 1335 (6.5) | 273 (2.2) | 566 (3.1) |
| Hypertension | 2260 (39.8) | 1900 (41.0) | 293 (50.0) | 1278 (48.8) | 16 325 (63.0) | 13 525 (66.0) | 3305 (26.3) | 5670 (31.3) |
| Diabetes | 770 (13.6) | 595 (12.8) | 68 (11.6) | 276 (10.5) | 5566 (21.5) | 4738 (23.1) | 1054 (8.4) | 1823 (10.1) |
| VTE (DVT/PE) | 1287 (22.7) | 704 (15.2) | 18 (3.1) | 9 (< 0.5) | 7282 (28.1) | 4381 (21.4) | 1335 (10.6) | 2189 (12.1) |
| Reduced renal functionf | ||||||||
| No | 3079 (54.2) | 2298 (49.6) | 215 (36.7) | 864 (33.0) | 888 (3.4) | 675 (3.3) | 12 499 (99.5) | 17 757 (98.2) |
| Yes | 872 (15.4) | 751 (16.2) | 57 (9.7) | 246 (9.4) | 3004 (11.6) | 3183 (15.5) | 58 (0.5) | 330 (1.8) |
| Unknown | 1729 (30.4) | 1587 (34.2) | 314 (53.6) | 1507 (57.6) | 22 022 (85.0) | 16 644 (81.2) | 0 (0.0) | 0 (0.0) |
| Time at risk, first episode of treatment,g days (used in the cohort analyses) | ||||||||
| Mean ± SD | 220.9 ± 303.0 | 296.6 ± 361.9 | 276.1 ± 209.0 | 253.5 ± 319.8 | 330.9 ± 376.7 | 503.0 ± 555.6 | 274.0 ± 235.0 | 562.0 ± 515.0 |
| Median (IQR) | 112.0 (42.0–224.0) | 190.0 (117.0–274.0) | 216.0 (126.0–399.0) | 200.0 (80.0–286.0) | 207.0 (102.0–412.0) | 260.0 (100.0–623.0) | 233.0 (145.0–286.0) | 333.0 (241.0–723.0) |
| Total follow-up time,h days (used in the nested case-control analysis) | ||||||||
| Mean ± SD | 841.0 ± 497.4 | 1209.0 ± 707.7 | 541.1 ± 267.2 | 1375.8 ± 606.4 | 1055.5 ± 486.5 | 1405.7 ± 570.1 | 1084.0 ± 485.0 | 1709.0 ± 650.0 |
| Median (IQR) | 770.5 (466.0–1196.0) | 1191.0 (606.0–1784.0) | 515.0 (390.0–672.0) | 1367.0 (931.0–1856.0) | 1033.0 (664.0–1434.0) | 1494.0 (969.0–1896.0) | 1086.0 (696.0–1445.0) | 1855.0 (1335.0–2218.0) |
Data are n (%) unless otherwise stated. Due to rounding, the sum of percentages do not always equal 100%
aPatients were categorized as naive or non-naive according to their previous use of any OAC other than their respective study drug any time before the start date
bThe number of outpatient visits or GP contacts is not assessable in GePaRD
cSwedish registers do not contain information on primary care visits. These values are, therefore, open-care visits, which include visits to hospital open clinics, emergency room visits without hospitalization and visits in specialized open care outside hospitals. Data on polypharmacy are not available from these records
dLifestyle information is not included in GePaRD. Obesity was identified by its diagnostic code. Although there is an ICD-10-GM code for heavy smoking, the data are not included here because it is expected that this information is only in the database if the person was treated for this condition
ePolypharmacy was defined as the number of medications (based on individual World Health Organization Anatomical Therapeutic Chemical codes) prescribed/dispensed in the 30 days (UK), 3 months (Netherlands) or 1 year (Germany) before the start date. It was not feasible to calculate polypharmacy in the Swedish health registers because information regarding the prescription of drugs were only available for specified medications of interest
fReduced renal function defined as eGFR < 50 mL/min/1.73 m2 (UK), eGFR < 60 mL/min/1.73 m2 (Netherlands), CKD stages 3–5 diagnostic codes and/or a procedure code for dialysis (Germany) or specific diagnostic codes and/or dispensation of drugs used in renal disease (Sweden)
gUntil treatment switching/discontinuation, occurrence of the outcome of interest, the end of the study period, death or withdrawal from the database
hUntil the end of the study period, death or withdrawal from the database
BMI, body mass index; CKD, chronic kidney disease; DVT, deep vein thrombosis; eGFR, estimated glomerular filtration rate; GePaRD, German Pharmacoepidemiological Research Database; GP, general practitioner; ICD-10-GM, International Statistical Classification of Diseases and Related Health Problems 10th Revision–German Modification; IQR, interquartile range; MI, myocardial infarction; NSAID, non-steroidal anti-inflammatory drug; OAC, oral anticoagulant; PE, pulmonary embolism; PPI, proton pump inhibitor; SD, standard deviation; VKA, vitamin K antagonist; VTE, venous thromboembolism
Incidence of bleeding events
The unadjusted IRs for the outcomes of interest are shown in Fig 1. In the UK, Germany and Sweden, intracranial bleeding IRs were similar for the rivaroxaban and VKA groups. The IR for intracranial bleeding was numerically higher for rivaroxaban users than VKA users in the Netherlands, but the 95% CIs overlapped. The urogenital bleeding IR was similar for rivaroxaban and VKA groups in the UK, the Netherlands and Sweden, although in Germany, it was higher in the rivaroxaban group than in the VKA group. In all four countries, gastrointestinal bleeding IRs were numerically higher in patients treated with rivaroxaban than in those treated with VKAs, but CIs overlapped in every comparison. The IR of other bleeding events was similar between treatment groups in the UK, the Netherlands and Sweden but lower in the rivaroxaban group than in the VKA group in Germany.
Fig 1. IRs per 100 person-years of safety outcomes associated with first use of rivaroxaban and VKAs for the treatment and secondary prevention of VTE.
CI, confidence interval; IR, incidence rate; Riva, rivaroxaban; VKA, vitamin K antagonist; VTE, venous thromboembolism.
Bleeding among specific high-risk subgroups of rivaroxaban users
Among rivaroxaban users in Germany and Sweden, the IRs of intracranial, gastrointestinal and other bleeding leading to hospitalization were all higher for the specific high-risk subgroups of patients than for the corresponding low-risk subgroups (patients with normal renal function, aged < 75 years or with no diabetes), although CIs were often overlapping (Table 2). For patients in Sweden who experienced urogenital bleeding, similar trends were found with respect to age and presence of diabetes. Conversely, urogenital bleeding IRs were higher in the low-risk subgroups than in the high-risk subgroups in Germany with respect to renal function, age and diabetes status. Overall, the results of the subgroup comparisons were more mixed in the UK and the Netherlands, where the numbers of events in rivaroxaban-treated patients were low (Table 2).
Table 2. IRs of intracranial, gastrointestinal, urogenital and other bleeding leading to hospitalization associated with first use of rivaroxaban in subgroups of specific interest.
| UK N = 5680 | Netherlandsa N = 586 | Germany N = 25 914 | Sweden N = 12 557 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Events, n | Person-years | IR (95% CI) | Events, n | Person-years | IR (95% CI) | Events, n | Person-years | IR (95% CI) | Events, n | Person-years | IR (95% CI) | |
| Intracranial bleeding | ||||||||||||
| All | 6 | 3435.5 | 0.17 (0.06–0.38) | 2 | 484 | 0.41 (0.04–1.2) | 67 | 23 448.5 | 0.29 (0.22–0.36) | 38 | 9420 | 0.40 (0.29–0.55) |
| Normal renal function | 5 | 1852.3 | 0.27 (0.09–0.63) | 1 | 209 | 0.5 (0.0–1.9) | 53 | 19 777.3 | 0.27 (0.20–0.35) | 37 | 9376 | 0.39 (0.29–0.54) |
| Impaired renal function | 1 | 603.7 | 0.17 (0.00–0.92) | 0 | 21 | N/A | 14 | 3671.2 | 0.38 (0.21–0.64) | 1 | 44 | 2.30 (0.32–16.30) |
| Age < 75 years | 4 | 2410.0 | 0.17 (0.05–0.42) | 1 | 365 | 0.3 (0.0–1.1) | 31 | 16 797.5 | 0.18 (0.13–0.26) | 22 | 7088 | 0.31 (0.20–0.47) |
| Age ≥ 75 years | 2 | 1025.3 | 0.20 (0.02–0.70) | 1 | 119 | 0.8 (0.0–3.4) | 36 | 6651.0 | 0.54 (0.38–0.75) | 16 | 2332 | 0.69 (0.42–1.12) |
| No diabetes | 6 | 2942.3 | 0.20 (0.07–0.44) | 2 | 433 | 0.5 (0.0–1.3) | 48 | 17 672.1 | 0.27 (0.20–0.36) | 31 | 8621 | 0.36 (0.25–0.51) |
| Type 1 or type 2 diabetes | 0 | 493.0 | 0.00 | 0 | 51 | N/A | 19 | 5776.4 | 0.33 (0.20–0.51) | 7 | 798 | 0.88 (0.42–1.84) |
| Gastrointestinal bleeding | ||||||||||||
| All | 20 | 3429.7 | 0.58 (0.36–0.90) | 3 | 483 | 0.62 (0.11–1.55) | 264 | 23 294.5 | 1.13 (1.00–1.28) | 86 | 9394 | 0.92 (0.74–1.13) |
| Normal renal function | 9b | 1850.9 | 0.49 (0.22–0.92) | 2 | 208 | 1.0 (0.1–2.8) | 177 | 19 690.7 | 0.90 (0.77–1.04) | 83 | 9351 | 0.89 (0.72–1.10) |
| Impaired renal function | 9b | 600.0 | 1.50 (0.69–2.85) | 0 | 21 | N/A | 87 | 3603.8 | 2.41 (1.93–2.98) | 3 | 43 | 6.94 (2.24–21.53) |
| Age < 75 years | 9 | 2407.2 | 0.37 (0.17–0.71) | 1 | 364 | 0.3 (0.0–1.1) | 106 | 16 730.7 | 0.63 (0.52–0.77) | 29 | 7081 | 0.41 (0.28–0.59) |
| Age ≥ 75 years | 11 | 1022.6 | 1.08 (0.54–1.92)) | 2 | 119 | 1.7 (0.1–4.9) | 158 | 6563.8 | 2.41 (2.05–2.81) | 57 | 2313 | 2.46 (1.90–3.19) |
| No diabetes | 16 | 2938.9 | 0.54 (0.31–0.88) | 3 | 434 | 0.7 (0.1–1.7) | 168 | 17 569.0 | 0.96 (0.82–1.11) | 72 | 8602 | 0.84 (0.66–1.05) |
| Type 1 or type 2 diabetes | 4 | 490.0 | 0.81 (0.22–2.09) | 0 | 49 | N/A | 96 | 5725.6 | 1.68 (1.36–2.05) | 14 | 791 | 1.77 (1.05–2.99) |
| Urogenital bleeding | ||||||||||||
| All | 10 | 3427.2 | 0.29 (0.14–0.54) | 3 | 483 | 0.62 (0.11–1.55) | 155 | 23 354.1 | 0.66 (0.56–0.78) | 31 | 9419 | 0.33 (0.23–0.47) |
| Normal renal function | 6c | 1844.6 | 0.33 (0.12–0.71) | 0 | 210 | N/A | 135 | 19 691.5 | 0.69 (0.57–0.81) | 31 | 9375 | 0.33 (0.23–0.47) |
| Impaired renal function | 2c | 603.7 | 0.33 (0.04–1.20) | 0 | 21 | N/A | 20 | 3662.6 | 0.55 (0.33–0.84) | 0 | 44 | N/A |
| Age < 75 years | 7 | 2401.8 | 0.29 (0.12–0.60) | 3 | 363 | 0.8 (0.1–2.1) | 124 | 16 721.9 | 0.74 (0.62–0.88) | 17 | 7089 | 0.24 (0.15–0.39) |
| Age ≥ 75 years | 3 | 1025.4 | 0.29 (0.06–0.86) | 0 | 120 | N/A | 31 | 6632.2 | 0.47 (0.32–0.66) | 14 | 2330 | 0.60 (0.36–1.01) |
| No diabetes | 8 | 2939.0 | 0.27 (0.12–0.54) | 3 | 433 | 0.7 (0.1–1.7) | 122 | 17 601.2 | 0.69 (0.58–0.83) | 22 | 8622 | 0.26 (0.17–0.39) |
| Type 1 or type 2 diabetes | 2 | 488.2 | 0.41 (0.05–1.48) | 0 | 49 | N/A | 33 | 5752.9 | 0.57 (0.39–0.81) | 9 | 797 | 1.13 (0.59–2.17) |
| Other bleeding | ||||||||||||
| All | 6 | 3433.2 | 0.17 (0.06–0.38) | 3 | 481 | 0.62 (0.11–1.55) | 103 | 23 402.7 | 0.44 (0.36–0.53) | 42 | 9414 | 0.45 (0.33–0.60) |
| Normal renal function | 4 | 1850.8 | 0.22 (0.06–0.55) | 1 | 209 | 0.5 (0.0–1.9) | 74 | 19 745.4 | 0.37 (0.29–0.47) | 41 | 9371 | 0.44 (0.32–0.59) |
| Impaired renal function | 2 | 603.0 | 0.33 (0.04–1.20) | 0 | 21 | N/A | 29 | 3657.3 | 0.79 (0.53–1.14) | 1 | 43 | 2.30 (0.32–1.63) |
| Age < 75 years | 3 | 2408.6 | 0.12 (0.03–0.36) | 2 | 362 | 0.6 (0.0–1.6) | 57 | 16 769.7 | 0.34 (0.26–0.44) | 22 | 7087 | 0.31 (0.20–0.47) |
| Age ≥ 75 years | 3 | 1024.6 | 0.29 (0.06–0.86) | 1 | 119 | 0.8 (0.0–3.4) | 46 | 6633.1 | 0.69 (0.51–0.93) | 20 | 2327 | 0.86 (0.55–1.33) |
| No diabetes | 5 | 2941.2 | 0.17 (0.06–0.40) | 3 | 430 | 0.7 (0.1–1.7) | 71 | 17 648.3 | 0.40 (0.31–0.51) | 38 | 8617 | 0.44 (0.32–0.61) |
| Type 1 or type 2 diabetes | 1 | 492.0 | 0.20 (0.01–1.13) | 0 | 51 | N/A | 32 | 5754.4 | 0.56 (0.38–0.79) | 4 | 797 | 0.50 (0.19–1.34) |
aPatients with unknown renal function were excluded from the analysis of data from the Netherlands, and, therefore, the number of patients from this country with normal renal function and impaired renal function do not always sum to the relevant total in the table
bTwo UK patients with gastrointestinal bleeding had missing or unknown eGFR values and, therefore, the number of patients from this country with normal renal function and impaired renal function does not sum to the total number of patients with gastrointestinal bleeding
cTwo UK patients with urogenital bleeding had missing or unknown eGFR values and, therefore, the number of patients from this country with normal renal function and impaired renal function does not sum to the total number of patients with urogenital bleeding
CI, confidence interval; eGFR, estimated glomerular filtration rate; IR, incidence rate; N/A, not available
Mortality
The IRs of all-cause mortality among rivaroxaban users increased with age and were higher in patients with renal impairment than in those without, and in patients with diabetes than in those without (Table 3). However, the CIs overlapped between treatment groups for all patient subgroups in the Netherlands, and when comparing by diabetes status in the UK.
Table 3. IRs of mortality per 100 person-years in first-time users of rivaroxaban (first episode of treatment).
| UK N = 5680 | Netherlandsa N = 586 | Germany N = 25 914 | Sweden N = 12 557 | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Events, n | Person-years | IR (95%CI) | Events, n | Person-years | IR (95% CI) | Events, n | Person-years | IR (95% CI) | Events, n | Person-years | IR (95% CI) | ||||
| All-cause mortality | |||||||||||||||
| All rivaroxaban | 139 | 3435.5 | 4.05 (3.40–4.78) | 9 | 485 | 1.9 (0.8–3.3) | 766 | 23 473.6 | 3.26 (3.04–3.50) | 304 | 9431 | 3.22 (2.88–3.61) | |||
| Analyses by renal function | |||||||||||||||
| Normal | 68a | 1852.0 | 3.67 (2.85–4.65) | 2b | 210 | 1.0 (0.1–2.8) | 473 | 19 794.9 | 2.39 (2.18–2.61) | 297 | 9387 | 3.16 (2.82–3.54) | |||
| Impaired | 42a | 604.0 | 6.95 (5.01–9.40) | 2b | 21 | 9.7 (0.8–28.3) | 293 | 3678.7 | 7.96 (7.08–8.93) | 7 | 44 | 16.06 (7.65–3.68) | |||
| Analyses by age | |||||||||||||||
| < 75 years | 42 | 2409.7 | 1.74 (1.26–2.36) | 3 | 365 | 0.8 (0.1–2.0) | 199 | 16 812.7 | 1.18 (1.02–1.36) | 106 | 7095 | 1.49 (1.23–1.81) | |||
| ≥ 75 years | 97 | 1025.7 | 9.46 (7.67–1.54) | 6 | 120 | 5.0 (1.8–9.9) | 567 | 6660.9 | 8.51 (7.83–9.24) | 198 | 2336 | 8.48 (7.37–9.74) | |||
| Analyses by diabetes status | |||||||||||||||
| No diabetes | 109 | 2942.4 | 3.70 (3.04–4.47) | 8 | 436 | 1.8 (0.8–3.4) | 470 | 17 690.7 | 2.66 (2.42–2.91) | 263 | 8632 | 3.05 (2.70–3.44) | |||
| Type 1 or type 2 diabetes | 30 | 493.0 | 6.08 (4.11–8.69) | 1 | 49 | 2.0 (0.0–8.1) | 296 | 5782.9 | 5.12 (4.55–5.74) | 41 | 799 | 5.13 (3.78–6.97) | |||
aTwenty-nine UK patients who died during the first episode of treatment had missing or unknown eGFR values and, therefore, the number of patients from this country with normal renal function and impaired renal function does not sum to the total number of patients who died
bFive Netherlands patients who died during the first episode of treatment had missing or unknown eGFR values and, therefore, the number of patients from this country with normal renal function and impaired renal function does not sum to the total number of patients who died
CI, confidence interval; eGFR, estimated glomerular filtration rate; IR incidence rate
Discussion
As part of the rivaroxaban PASS programme [10], data from routine clinical practice in four European countries were used to assess risk of bleeding in patients treated for VTE with rivaroxaban or VKA. The safety profile of rivaroxaban in the PASS was consistent with previously reported profiles; no new safety concerns were identified. Unadjusted IRs of safety outcomes among rivaroxaban users for the treatment and secondary prevention of VTE were generally in line with the cumulative incidences reported for the EINSTEIN-DVT and EINSTEIN-PE RCTs and for other real-world studies [11, 12, 18–20].
This PASS was not designed for direct statistical IR comparisons between treatment groups and, because most CIs were overlapping, it was difficult to ascertain trends for consistent differences across countries. Despite this, although both RCTs and real-world data generally showed a similar risk of major bleeding associated with rivaroxaban use compared with VKA use for intracranial and urogenital bleeding [21–25], there is some evidence in the PASS cohorts of increased IRs of gastrointestinal bleeds leading to hospitalization in patients receiving rivaroxaban than in patients receiving VKAs. This is consistent with a meta-analysis that showed increased risk of gastrointestinal bleeding following treatment with direct oral anticoagulants compared with warfarin across indications [26] and also in line with studies on AF patients [27]. However, this was not consistent with other studies in VTE that favoured rivaroxaban over VKAs with respect to gastrointestinal bleeding [22, 23].
Many of the reported observational studies in VTE have follow-up periods of 6 months or less, do not separate major bleeding by site, or do not report IRs by patient-years, thereby making comparisons with the findings reported in the present study difficult. A pooled analysis of the prospective, non-interventional XALIA and XALIA-LEA studies across 36 countries observed patients for more than 6 months and reported unadjusted IRs (95% CI) of any major bleeding of 1.74 (1.24–2.38) per 100 patient-years for rivaroxaban and 3.9 (3.03–5.05) per 100 patient-years for VKAs [20]. In the DRESDEN registry, the IR (95% CI) of major bleeding was 4.1 (2.5–6.4) per 100 patient-years for patients with VTE (median treatment duration, 274 days) [18]. In comparison, the rate of major bleeding in a UK cohort observed for only 12 weeks was higher than in the aforementioned studies, as the cumulative IR (95% CI) of major bleeding with rivaroxaban for the treatment and secondary prevention of VTE was 8.3 (5.3–12.5) [19]. Interestingly, as in previous studies, the results from the PASS also showed IR variations by country. There are likely to be numerous reasons for this observation, including the diversity of patient characteristics from each country and the known differences between healthcare systems. For example, differences in the speed in which each healthcare system recommended the use of rivaroxaban for an indication of VTE inevitably resulted in discrepancies in rivaroxaban uptake compared with VKA uptake in each country.
Consistent with known risk factors and clinical trials [3, 9], our data showed that IRs of intracranial, gastrointestinal and other bleeding events were higher in patients aged 75 years or older. Older age was also found to be an independent risk factor for major bleeding in a study of Swedish national registries (normalized hazard ratio [95% CI]: 1.38 [1.27–1.50]) [28].
In the PASS, the IRs of all-cause mortality were higher in rivaroxaban users aged 75 years or older and in those with impaired renal function or diabetes. The IR (95% CI) of all-cause mortality in this study was nominally higher than that in the pooled analysis of XALIA and XALIA-LEA, which was 1.83 (1.31–2.48) events per 100 patient-years, but many factors could contribute to this difference [20].
Interpretations of the findings must consider the study limitations, some of which are inherent in observational data. The healthcare databases used existed before the study and primarily contained data collected routinely in clinical practice. The objectives, methodologies and codes used to ascertain each outcome and variable were validated and harmonized to the extent feasible between the studies to facilitate descriptive comparisons between countries. However, because of residual differences in the data sources and irreconcilable differences in terms of data availability, no pooled analyses of the studies or statistical comparisons were performed. Nevertheless, although point estimates varied between countries, the direction of trends and magnitudes of effect were broadly similar, except when noted.
Although it is of interest to interpret the results of this PASS in the context of the pivotal RCT findings, neither the background information nor the outcome events are captured in the same manner by these two distinct study approaches. Fundamentally, the patient populations are different; for example, patients were older in the PASS (mean age, 60–64 years) than in the EINSTEIN-DVT and EINSTEIN-PE studies (pooled mean age, 57 years) [9]. However, it is axiomatic that differences in patient characteristics are observed between patients from a real-world study compared with those who have met the strict selection criteria of an RCT. Real-world studies are, therefore, critical in informing benefit–risk assessments in under-represented groups.
Conclusions
This PASS provides data from routine clinical practice that support the safety profile of rivaroxaban, established from previous RCTs and other post-authorization studies. These findings provide further data to support physicians’ decision-making when prescribing rivaroxaban for VTE.
Supporting information
(DOCX)
(DOCX)
(DOCX)
Acknowledgments
Editorial support initial versions of the manuscript was provided by Oxford PharmaGenesis, Oxford, UK.
Data Availability
All relevant data are within the manuscript and its supporting information files. The study uses individual patient electronic health record from different countries, the data on the respective databases cannot be shared publicly and are available for scientific researchers who meet relevant criteria and through the appropriate research ethics committees and regulatory authorities. For the UK, the IQVIA Medical Research Data(IMRD) incorporating THIN, protocol approval is received from the Scientific Research Committee (SRC 14-018) (contact via info@the-health-improvement-network.com). For Swedish Nationwide Health Registries the Stockholm Ethical Review Board and the Swedish Board of Health and Welfare (contact via: lupop@ed.lu.es) In Germany, in accordance with German data protection regulations, access to the data from The German Pharmacoepidemiological Research Database (GePaRD) is granted only to employees of the Leibniz Institute for Prevention Research and Epidemiology – BIPS on the premises of the institute and in the context of approved research projects. Third parties may only access the data in cooperation with BIPS and after signing an agreement for guest researchers, (contact via: kontakt(at)leibniz-bips.de). Data from the Netherlands are only available on reasonable request, in cooperation with the PHARMO Institute for Drug Outcomes Research, and after approval of the Compliance Committee of STIZON, (contact via: www.pharmo.com; pharmo@pharmo.nl).
Funding Statement
This PASS was requested by the European Medicines Agency, financed by Bayer AG and performed in line with the ENCePP Code of Conduct. Initial versions of the manuscript received editorial support from Oxford PharmaGenesis, Oxford, UK, with funding from Bayer AG. The funders had no role in study design, data collection and analysis, or decision to publish the manuscript.
References
- 1.Cohen AT, Agnelli G, Anderson FA, Arcelus JI, Bergqvist D, Brecht JG, et al. Venous thromboembolism (VTE) in Europe. The number of VTE events and associated morbidity and mortality. Thromb Haemost. 2007;98(4):756–64. [DOI] [PubMed] [Google Scholar]
- 2.Anderson FA Jr, Spencer FA. Risk factors for venous thromboembolism. Circulation. 2003;107(23_suppl_1):I–9. [DOI] [PubMed] [Google Scholar]
- 3.Konstantinides SV, Meyer G, Becattini C, Bueno H, Geersing GJ, Harjola VP, et al. 2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS). Eur Heart J. 2020;41(4):543–603. doi: 10.1093/eurheartj/ehz405 [DOI] [PubMed] [Google Scholar]
- 4.Jha AK, Larizgoitia I, Audera-Lopez C, Prasopa-Plaizier N, Waters H, Bates DW. The global burden of unsafe medical care: analytic modelling of observational studies. BMJ Qual Saf. 2013;22(10):809–15. doi: 10.1136/bmjqs-2012-001748 [DOI] [PubMed] [Google Scholar]
- 5.Khan F, Rahman A, Carrier M, Kearon C, Weitz JI, Schulman S, et al. Long term risk of symptomatic recurrent venous thromboembolism after discontinuation of anticoagulant treatment for first unprovoked venous thromboembolism event: systematic review and meta-analysis. BMJ. 2019;366:l4363. doi: 10.1136/bmj.l4363 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ortel TL, Neumann I, Ageno W, Beyth R, Clark NP, Cuker A, et al. American Society of Hematology 2020 guidelines for management of venous thromboembolism: treatment of deep vein thrombosis and pulmonary embolism. Blood Adv. 2020;4(19):4693–738. doi: 10.1182/bloodadvances.2020001830 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bayer AG. Summary of Product Characteristics, Xarelto. 2018. https://www.ema.europa.eu/en/documents/product-information/xarelto-epar-product-information_en.pdf. Accessed 20 Feb 2023 [Google Scholar]
- 8.Janssen Pharmaceuticals, Inc. Prescribing Information, XARELTO®. 2020. https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/202439s031,022406s035lbl.pdf. Accessed Feb 20 2023 [Google Scholar]
- 9.Prins MH, Lensing AW, Bauersachs R, van Bellen B, Bounameaux H, Brighton TA, et al. Oral rivaroxaban versus standard therapy for the treatment of symptomatic venous thromboembolism: a pooled analysis of the EINSTEIN-DVT and PE randomized studies. Thromb J. 2013;11(1):21. doi: 10.1186/1477-9560-11-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Garcia-Rodriguez LA, Wallander MA, Friberg L, Ruigomez A, Schink T, Bezemer I, et al. Rationale and design of a European epidemiological post-authorization safety study (PASS) program: rivaroxaban use in routine clinical practice. Expert Opin Drug Saf. 2020;19(11):1513–20. doi: 10.1080/14740338.2020.1798928 [DOI] [PubMed] [Google Scholar]
- 11.Einstein Investigators, Bauersachs R, Berkowitz SD, Brenner B, Buller HR, Decousus H, et al. Oral rivaroxaban for symptomatic venous thromboembolism. N Engl J Med. 2010;363(26):2499–510. doi: 10.1056/NEJMoa1007903 [DOI] [PubMed] [Google Scholar]
- 12.Investigators Einstein-PE, Buller HR Prins MH, Lensin AW, Decousus H, Jacobson BF, et al. Oral rivaroxaban for the treatment of symptomatic pulmonary embolism. N Engl J Med. 2012;366(14):1287–97. doi: 10.1056/NEJMoa1113572 [DOI] [PubMed] [Google Scholar]
- 13.Angelini DE, Radivoyevitch T, McCrae KR, Khorana AA. Bleeding incidence and risk factors among cancer patients treated with anticoagulation. Am J Hematol. 2019;94(7):780–5. doi: 10.1002/ajh.25494 [DOI] [PubMed] [Google Scholar]
- 14.Prandoni P, Lensing AW, Piccioli A, Bernardi E, Simioni P, Girolami B, et al. Recurrent venous thromboembolism and bleeding complications during anticoagulant treatment in patients with cancer and venous thrombosis. Blood. 2002;100(10):3484–8. doi: 10.1182/blood-2002-01-0108 [DOI] [PubMed] [Google Scholar]
- 15.Hutten BA, Prins MH, Gent M, Ginsberg J, Tijssen JG, Büller HR. Incidence of recurrent thromboembolic and bleeding complications among patients with venous thromboembolism in relation to both malignancy and achieved international normalized ratio: a retrospective analysis. J Clin Oncol. 2000;18(17):3078–83. doi: 10.1200/JCO.2000.18.17.3078 [DOI] [PubMed] [Google Scholar]
- 16.Ruigómez A, Brobert G, Suzart-Woischnik K, García-Rodríguez LA. Ascertainment and validation of major bleeding events in a primary care database. Pharmacoepidemiol Drug Saf. 2019;28(2):148–55. doi: 10.1002/pds.4580 [DOI] [PubMed] [Google Scholar]
- 17.Ruigómez A, Brobert G, Vora P, García Rodríguez LA. Validation of venous thromboembolism diagnoses in patients receiving rivaroxaban or warfarin in The Health Improvement Network. Pharmacoepidemiol Drug Saf. 2021;30(2):229–36. doi: 10.1002/pds.5146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Beyer-Westendorf J, Forster K, Pannach S, Ebertz F, Gelbricht V, Thieme C, et al. Rates, management, and outcome of rivaroxaban bleeding in daily care: results from the Dresden NOAC registry. Blood. 2014;124(6):955–62. doi: 10.1182/blood-2014-03-563577 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Evans A, Davies M, Osborne V, Roy D, Shakir S. Evaluation of the incidence of bleeding in patients prescribed rivaroxaban for the treatment and prevention of deep vein thrombosis and pulmonary embolism in UK secondary care: an observational cohort study. BMJ Open. 2020;10(11):e038102. doi: 10.1136/bmjopen-2020-038102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Haas S, Mantovani LG, Kreutz R, Monje D, Schneider J, Zell ER, et al. Anticoagulant treatment for venous thromboembolism: A pooled analysis and additional results of the XALIA and XALIA-LEA noninterventional studies. Res Pract Thromb Haemost. 2021;5(3):426–38. doi: 10.1002/rth2.12489 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Coleman CI, Bunz TJ, Turpie AGG. Effectiveness and safety of rivaroxaban versus warfarin for treatment and prevention of recurrence of venous thromboembolism. Thromb Haemost. 2017;117(10):1841–7. doi: 10.1160/TH17-03-0210 [DOI] [PubMed] [Google Scholar]
- 22.Coleman CI, Peacock WF, Bunz TJ, Beyer-Westendorf J. Effectiveness and safety of rivaroxaban versus warfarin in patients with unprovoked venous thromboembolism: A propensity-score weighted administrative claims cohort study. Thromb Res. 2018;168:31–6. doi: 10.1016/j.thromres.2018.05.031 [DOI] [PubMed] [Google Scholar]
- 23.Kohn CG, Bunz TJ, Beyer-Westendorf J, Coleman CI. Comparative risk of major bleeding with rivaroxaban and warfarin: population-based cohort study of unprovoked venous thromboembolism. Eur J Haematol. 2019;102(2):143–9. doi: 10.1111/ejh.13185 [DOI] [PubMed] [Google Scholar]
- 24.Sindet-Pedersen C, Langtved Pallisgaard J, Staerk L, Gerds TA, Fosbøl EL, Torp-Pedersen C, et al. Comparative safety and effectiveness of rivaroxaban versus VKAs in patients with venous thromboembolism. A Danish nationwide registry-based study. Thromb Haemost. 2017;117(6):1182–91. doi: 10.1160/TH16-10-0745 [DOI] [PubMed] [Google Scholar]
- 25.Kucher N, Aujesky D, Beer JH, Mazzolai L, Baldi T, Banyai M, et al. Rivaroxaban for the treatment of venous thromboembolism. The SWIss Venous ThromboEmbolism Registry (SWIVTER). Thromb Haemost. 2016;116(3):472–9. doi: 10.1160/TH16-03-0209 [DOI] [PubMed] [Google Scholar]
- 26.Holster IL, Valkhoff VE, Kuipers EJ, Tjwa E. New oral anticoagulants increase risk for gastrointestinal bleeding: a systematic review and meta-analysis. Gastroenterology. 2013;145(1):105–12.e15. doi: 10.1053/j.gastro.2013.02.041 [DOI] [PubMed] [Google Scholar]
- 27.Ballestri S, Romagnoli E, Arioli D, Coluccion V,Marrazzo A, Athasasiou A et al. Risk and management of bleeding complications with direct oral anticoagulants in patients with atrial fibrillation and venous thromboembolism: a narrative review. Adv Ther. 2023;40:41–66 doi: 10.1007/s12325-022-02333-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wallvik N, Renlund H, Själander A. Bleeding risk in patients with venous thromboembolic events treated with new oral anticoagulants. J Thromb Thrombolysis. 2020:315–23. doi: 10.1007/s11239-020-02319-w [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX)
(DOCX)
(DOCX)
Data Availability Statement
All relevant data are within the manuscript and its supporting information files. The study uses individual patient electronic health record from different countries, the data on the respective databases cannot be shared publicly and are available for scientific researchers who meet relevant criteria and through the appropriate research ethics committees and regulatory authorities. For the UK, the IQVIA Medical Research Data(IMRD) incorporating THIN, protocol approval is received from the Scientific Research Committee (SRC 14-018) (contact via info@the-health-improvement-network.com). For Swedish Nationwide Health Registries the Stockholm Ethical Review Board and the Swedish Board of Health and Welfare (contact via: lupop@ed.lu.es) In Germany, in accordance with German data protection regulations, access to the data from The German Pharmacoepidemiological Research Database (GePaRD) is granted only to employees of the Leibniz Institute for Prevention Research and Epidemiology – BIPS on the premises of the institute and in the context of approved research projects. Third parties may only access the data in cooperation with BIPS and after signing an agreement for guest researchers, (contact via: kontakt(at)leibniz-bips.de). Data from the Netherlands are only available on reasonable request, in cooperation with the PHARMO Institute for Drug Outcomes Research, and after approval of the Compliance Committee of STIZON, (contact via: www.pharmo.com; pharmo@pharmo.nl).