Thrombosis Research 183 (2019) 28–32

Contents lists available at ScienceDirect

Thrombosis Research
journal homepage: www.elsevier.com/locate/thromres

Full Length Article

Very elderly patients with venous thromboembolism on oral anticoagulation T

with VKAs or DOACs: Results from the prospective multicenter START2-
Register Study
⁎
Daniela Polia, , Emilia Antonuccib, Lorenza Bertùc, Elisa Vigninid, Lucia Ruoccoe,
Daniela Mastroiacovof, Carmelo Paparog, Daniele Pastorih, Sophie Testai, Walter Agenoc,
Gualtiero Palaretib, coordinator of START2 Register
a
Center for Atherothrombotic Diseases, Azienda Ospedaliero-Universitaria Careggi, Firenze, Italy
b
Arianna Anticoagulazione Foundation, Bologna, Italy
c
Department of Medicine and Surgery, University of Insubria, Varese, Italy
d
Department of Experimental and Clinical Medicine, University of Florence, Firenze, Italy
e
Thrombosis Center, U.O. Analisi Chimico-Cliniche, Azienda Ospedaliero-Universitaria Pisana, Pisa, Italy
f
Department of Angiology, Ospedale SS. Filippo e Nicola, Avezzano, L'Aquila, Italy
g
Thrombosis Center, Ospedale Maggiore, Chieri, Torino, Italy
h
Atherothrombosis Centre, Department of Internal Medicine, Sapienza University, Rome, Italy
i
Haemostasis and Thrombosis Centre, Hospital of Cremona, Cremona, Italy

A R T I C LE I N FO A B S T R A C T

Keywords: Introduction: Few data are available on the safety of anticoagulation in very elderly patients treated with
Bleeding Vitamin K antagonists (VKAs) and direct oral anticoagulants (DOACs) for venous thromboembolism (VTE).
Direct oral anticoagulant Methods: We carried out a prospective cohort study on VTE patients aged ≥85 years enrolled in the Survey on
Elderly anticoagulaTed pAtients RegisTer (START2-Register) on treatment with VKAs or DOACs, with the aim to
Venous thromboembolism
evaluate mortality, bleeding and thrombotic rates (venous and arterial).
Vitamin K antagonist
Results: We enrolled 272 patients, 58.7% on VKA and 41.3% on DOACs. Baseline characteristics were similar
between treatment groups, with a higher prevalence of renal failure in VKAs patients and of a history of bleeding
and previous stroke/TIA in DOACs patients. During follow-up of 429 patient-years, 15 major and non-major
clinically relevant bleedings were recorded (rate 3.5 × 100 pt-yrs), 5 were major bleeds (rate 1.2 × 100 pt-yrs),
1 in a patient on aspirin (rate 4.3 × 100 pt-yrs). Bleeding rate was higher in patients on DOACs (crude HR 4.7;
95%CI 1.5–15.01). Eight thrombotic events were recorded (rate 1.9 × 100 pt-yrs), 3 recurrent VTE and 5 stroke/
TIA. Overall, the incidence of thrombotic events was higher in DOACs patients (crude HR 4.5; 95% CI 1.5; 13.3).
The rate of recurrent VTE was similar in the two group. Mortality rate was significantly lower in DOACs patients
(crude HR 0.30; 95% CI 0.1;0.9).
Conclusion: A higher bleeding risk was found in very elderly VTE patients on DOACs despite the wide use of low-
dosages. Similarly a higher thrombotic risk was found while the incidence of recurrent VTE was low and similar
between the groups. Mortality rate were significantly lower in DOACs patients.

1. Introduction pulmonary hypertension. A first episode of VTE confers a higher risk for
recurrences, therefore VTE must be considered a chronic disease.
The term “venous thromboembolism” (VTE) includes deep vein Oral anticoagulation either with vitamin K antagonists (VKAs) or
thrombosis (DVT) of the lower limbs and/or pulmonary embolism (PE). direct oral anticoagulants (DOACs) has a proven efficacy in the short
VTE is a common disease, especially in the elderly, and is associated and long-term prevention of VTE recurrences. However, the long-term
with increased risk of mortality, hospitalization and chronic compli- safety of anticoagulant treatment remains a matter of concern, in par-
cations such as post-thrombotic syndrome or chronic thromboembolic ticular for elderly patients. No randomized-controlled trials (RCTs) of

⁎
Corresponding author at: Centro Trombosi, Azienda Ospedaliero Universitaria Careggi, Viale Morgagni, 85–50134 Firenze, Italy.
E-mail address: polida@aou-careggi.toscana.it (D. Poli).

https://doi.org/10.1016/j.thromres.2019.08.024
Received 17 May 2019; Received in revised form 23 August 2019; Accepted 24 August 2019
Available online 26 August 2019
0049-3848/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
D. Poli, et al. Thrombosis Research 183 (2019) 28–32

DOACs versus VKA were specifically designed to address this issue in a quarterly, by phone call or ambulatory visit. An ambulatory follow-up
population of elderly patients. A meta-analysis of clinical trials showed visit is mandatory at least annually. In the presence of severe dementia
that in patients aged 75 years or more, DOACs were more effective in or frequent falls or bed rest the patients were defined frail. Creatinine
reducing VTE recurrences and safer than warfarin [1,2]. However, in clearance was calculated by the Cockroft-Gault formula [7]. Patients
these RCTs elderly patients represented only about 12–15% of the are stratified for baseline bleeding risk evaluation according to HAS-
whole study populations, with the exception of the Recover II study BLED score [8].
(20%) and the Resonate study (30%) [3]. Major endpoints of the study were bleeding events, VTE recurrence,
Therefore, additional evidence is needed on the safety and effec- and death for all causes. Major bleeding (MB) were defined as re-
tiveness of oral anticoagulant therapy (VKAs or DOACs) in very elderly commended by the International Society on Thrombosis and
patients with acute VTE and for those who require extended treatment. Haemostasis [9]. Clinically relevant non-major bleeding (CRNMB) are
Moreover, it is worth collecting data from real-world patients about the defined as any overt bleeding requiring a medical intervention (hospi-
use of low-dose DOACs in this setting. talization, surgery or interventional procedure, further diagnostic
We have carried out a multicenter, prospective cohort study on imaging, laboratory test or specialist evaluation) and/or treatment
patients aged ≥85 years with a VTE event enrolled in the Survey on discontinuation, and not meeting any of the criteria for major bleeding
anticoagulaTed pAtients RegisTer (START2-Register) who started a [10].
treatment with either VKAs or DOACs to evaluate the type and duration Thromboembolic complications, such as new or recurrent DVT or PE
of anticoagulant therapy and the incidence of adverse events. episodes, stroke, transient ischemic attack (TIA), peripheral embolism,
and acute myocardial infarction (AMI) are recorded. If not specified
2. Methods differently in relation to specific study designs, thrombotic complica-
tions are adjudicated by the local investigators, based on clinical signs
The START2-Register is an observational, multicenter, prospective and symptoms combined with objectively confirmed diagnostic radi-
cohort study that includes adults (> 18 years) who start antic- ology and laboratory tests (color-Doppler ultrasound investigation,
oagulation therapy, whatever the clinical indication for the therapy, the magnetic resonance imaging, computed tomography, electro-
drug and dosage used [4]. The aim of the START2-Register is to collect cardiography, laboratory markers). Death for all causes were con-
data on effectiveness and safety of anticoagulant treatments, on de- sidered; in particular death was defined as consequence of bleeding or
terminants of adverse events in anticoagulated patients, as well as on of thrombotic events or not related to anticoagulation in all the other
their quality of life and compliance to treatment. The registry has been cases.
approved on October 2011 (N = 142/2010/0/0ss) by the Ethical
Committee of the Institution of the Coordinating Member (AziendaOs- 2.1. Statistical analysis
pedaliero-Universitaria, Policlinico S. Orsola-Malpighi, Bologna, Italy),
all patients gave their written informed consent before enrollment. The Data were described as the mean value and standard deviation (SD)
study is registered in ClinicalTrials.gov Identifier: NCT02219984; it is for continuous variables and proportions for categorical variables.
ongoing and actively recruiting. Here we present the results of the co- Differences between continuous values were assessed using the un-
hort of patients with venous thromboembolism (VTE) who started the paired t-test, categorical variables were compared by the Chi-square
anticoagulation at the age ≥85 years. For the purpose of the present test or Fisher exact test as appropriate.
analysis, data were collected from January 2012 to April 2018. VTE The median and interquartile range (IQR) follow-up time were
events are considered as unprovoked if they are not temporally asso- calculated and the median test applied to test difference between
ciated with a major risk factor within 3 months of diagnosis (surgery groups.
with general or spinal anesthesia, lower limb fracture, casting or no The incidences of death, thrombotic accidents, and bleeding events
weight bearing for ≥3 days, bed-bound for 3 days due to acute illness), were calculated by dividing the number of events by person time at risk.
or are associated with a minor risk factor [5]. The incidence rate ratio and together with the 95% confidence interval
All participating centers are asked to consecutively include patients (95%CI) were calculated.
who start anticoagulant treatment, for any indication and with any The crude hazard ratios (HRs) with the 95%CI, and a multivariate
available drug, if this is planned to last for at least 3 months. Patients hazard model was calculated for the three outcomes. All analysis was
with life-expectancy < 6 months, or not residents in the participant carried out using SAS statistical package (Version 9.4 for Windows. SAS
region, or planning to leave in the next six months after enrolment are Institute Inc. Cary NC).
not eligible, as well as patients already enrolled in phase II or III clinical
studies. Patients were followed-up during anticoagulant treatment, and 3. Results
follow-up was stopped when the treatment was withdrawn. Participants
are required to enroll their patients consecutively, without any a priori The present study analyzed 272 patients who started antic-
exclusion criteria other than life-expectancy or geographical in- oagulation at the age ≥85 years, and who were included in the
accessibility. Definition of the time-frame for enrolment (e. g. one week START2-Register during the interval time February 2012–April 2018
every month, or the first month of the year) is left at each participant's for the occurrence of a VTE episode. The baseline characteristics of the
discretion, as long as it provides a random enrolment of patients. whole population are detailed in Table 1. The mean age of patients was
Baseline patient's clinical features are recorded by participants on web- 88.1 years, and patients on DOACs were significantly older than pa-
based case report forms (CRF) and include: demographic and clinical tients on VKAs. Sixty-nine patients (25.4%) were aged ≥90 years at the
characteristics of patients, clinical indication for treatment, associated beginning of treatment. Patients were followed for a total period of 429
risk factors for thromboembolic complications or bleeding occurring patient-years; 7 (2.6%) patients were lost at follow-up, 52 (19.0%)
during treatment, laboratory routine data, type of anticoagulant drug patients stopped treatment after the established period of treatment, 12
used and dose (or expected therapeutic range), use of concomitant (4.4%) stopped treatment for the onset of a contraindication; 32
drugs. For patients treated with VKA, all INR controls, the subsequent (11.7%) patients still on treatment were no longer followed by the
dosing prescriptions and information at each visit about possible clin- participating centre.
ical events and changes in the medical history are automatically cap- The anticoagulant treatment was warfarin in 156 patients (57.3%)
tured via informatics. All INRs were recorded and time in therapeutic and direct oral anticoagulants (DOACs) in 116 patients (42.7%), and 15
range (TTR) [6] of the last 6 months of treatment was reported. Parti- patients (5.5%) were also on treatment with antiplatelet drugs (9 with
cipants are required to regularly follow-up all enrolled patients at least low-dose aspirin, and 6 with clopidogrel), as reported in Table 2.

29
D. Poli, et al. Thrombosis Research 183 (2019) 28–32

Table 1 Table 3
Clinical characteristics of patients. Characteristics of VTE index event.
All patients VKA DOACs p Value All patients VKA DOACs p Value
(N = 272) (N = 156) (N = 116; (N = 272) (N = 156) (N = 116)
naive 93,
80.2%) N% N% N%

N% N% N% Site of index event

DVT 148 (54.4) 82 (52.6) 66 (56.9) 0.5
Sex- female 178 (65.4) 99 (63.5) 79 (68.1) 0.4 DVT/PE 57 (21.0) 32 (20.5) 25 (21.6) 0.9
Age (years) - mean (SD) 88.1 (3.0) 87.6 (2.3) 88.8 (3.7) 0.001 PE 67 (24.6) 42 (26.9) 25 (21.6) 0.3
Weight < 60 kg 65 (23.9) 42 (26.9) 23 (19.8) 0.17 Nature of index event (270/272
Follow-up (years) 429 312 117 patients)
Mean follow-up (months) 19.1 (16.8) 24.3 (19.0) 12.2 (9.6) 0.01 Unprovoked 217 (80.3) 126 (80.8) 91 (80.2) 0.2
(SD) Secondary to transient risk 53 (19.7) 30 (19.2) 23 (19.8) 1.0
Co-morbidity factor
Creatinine clearance 34 (12.5) 28 (17.9) 6 (5.2) 0.001 Recurrent VTE 28 (10.4) 14 (9.0) 14 (12.1) 0.4
< 30 mL/min
Creatinine clearance 150 (55.1) 83 (53.2) 67 (57.8) 0.46
30–50 mL/min
(54.4%), and VTE was unprovoked in 217 patients (79.8%), with a si-
Hemoglobin < 10 g/dL 12 (4.7) 7 (4.5) 5 (4.3) 1.0
Platelet count < 100,000 3 (1.1) 2 (1.3) 1 (0.9) 1.0
milar distribution between patients on warfarin and patients on DOACs
Chronic disease 16 (5.9) 10 (6.4) 6 (5.2) 0.8 (Table 3).
Active cancer 11 (4.0) 9 (5.8) 2 (1.7) 0.2 As shown in Table 4, during follow-up 15 major and CRNM bleeding
Diabetes mellitus 36 (13.0) 16 (10.3) 20 (17.2) 0.1 events were recorded in the whole study population (rate 3.5 × 100 pt-
Hypertension 191 (70.2) 114 (73.1) 77 (66.4) 0.2
yrs); 5 events were major bleeding (rate 1.2 × 100 pt-yrs), one of them
Previous stroke/TIA 27 (9.9) 10 (6.4) 17 (14.7) 0.04
Previous bleeding 11 (4.0) 3 (1.9) 8 (6.9) 0.06 (gastrointestinal in a patient on warfarin) was fatal; 6/15 bleeding
Coronary artery disease 29 (10.7) 19 (12.2) 10 (8.6) 0.4 events occurred among patients aged ≥90 years, all on DOACs treat-
Heart failure 18 (6.6) 11 (7.1) 7 (6.0) 0.8 ment. One major bleeding was recorded in patients concomitantly
POAD 18 (6.6) 9 (5.8) 9 (7.8) 0.6
treated with aspirin (rate 4.3 × 100 pt-yrs). Bleeding rates were higher
COPD 38 (14.0) 24 (15.4) 14 (12.1) 0.5
Frail subjectsa 50 (18.4) 31 (19.9) 19 (16.4) 0.5
in patients on DOACs than in patients on VKAs (HR 4.0; 95%CI
Bleeding risk stratification 1.3–13.6; p = 0.01). These results are confirmed also when the analysis
scores was limited to the first year of follow-up (Table 4). The difference was
HASBLED - mean (SD) 2.0 (0.7) 2.05 (0.7) 2.0 (0.8) NS confirmed at the multivariate model; 4 of 9 bleeding events in patients
a treated with DOACs occurred on full dose treatment.
Patient with dementia or bed rest or prone to fall.
Overall, 8 thrombotic events were recorded during treatment (rate
1.9 × 100 pt-yrs); 2 were fatal (one each in patients on DOACs and on
Table 2
warfarin) (Table 4). The rate of thrombotic events was significantly
Type of treatment and co-medications.
higher in patients on DOACs than in patients on VKAs (HR 4.5; 95% CI
All patients VKA DOACs p Value 1.5–13.3); this difference was confirmed at the multivariate model and
(N = 272) (N = 156) (N = 116)
when the analysis was limited to the first year. Only 3 of these events
N% N% N% were recurrences of VTE and no difference was detected between the 2
groups.
Type of anticoagulant drug Mortality rates were lower in patients on DOACs than in those on
Warfarin 156 (100)
warfarin, with a HR of 0.30 (95% CI 0.1;0.9), confirmed at the multi-
Apixaban 43 (37.1)
Dabigatran 5 (4.3) variate model (Table 4). These results are confirmed also when the
Edoxaban 7 (6.0) analysis was limited to the first year of follow-up (HR 0.19, 95% CI
Rivaroxaban 61 (52.6) 0.1–0.6).
Low-dose DOACs 52 (44.8)
Co-medications
Antiplatelet drugs 15 (5.5) 9 (5.8) 6 (5.2) 1.0 4. Discussion
Proton pump inhibitor 127 (46.7) 79 (50.6) 48 (41.4) 0.8
Statins 43 (15.8) 19 (12.2) 24 (20.7) 0.1
Psychotropic drugs 48 (17.6) 30 (19.2) 18 (15.5) 0.6 In this observational, prospective cohort study of very old patients
on anticoagulant treatment for VTE, the bleeding risk was low despite
the mean age of the population, with nearly 25% of patients older than
Among DAOC patients 93, (80.2%) were naïve to treatment. Patients on 90 years at enrollment. The median duration of treatment was
warfarin showed a median time in therapeutic range (TTR) of 68% 19 months, thus showing that elderly patients enrolled in this study are
(interquartile range 54–79%); 52/116 (44.8%) patients on DOACs were maintained on anticoagulant treatment for a long period of time despite
on treatment with the low dosage available for the treatment of acute clinical guidelines suggesting stopping anticoagulation after 3 months
VTE for each drug. in patients at high risk of bleeding, such as very old patients.
The prevalence of patients with severe renal failure (creatinine When we looked at the type of anticoagulant drug used, we found
clearance < 30 mL/min) was significantly higher in patients on war- that patients on DOACs had higher rates of bleeding and also of
farin with respect to patients on DOACs. Conversely, the prevalence of thrombotic complications with respect to patients on warfarin; how-
history of stroke/TIA and history of bleeding was significantly higher in ever, their mortality rate was markedly reduced.
DOACs patients (Table 1). All other clinical characteristics were similar The two groups had different baseline characteristics, a finding that
between patients on warfarin and patients on DOACs (Table 1). Con- is explained by the observational nature of the study. However, these
comitant use of antiplatelet agents was also similar between the 2 differences did not reach statistical significance in the majority of cases,
treatment groups (Table 2). likely due to the relatively low number of enrolled patients. Patients on
The index event was deep vein thrombosis (DVT) in 148 patients DOACs were significantly older and with a more prevalent history of
previous bleeding or stroke/TIA events than patients on warfarin. On

30
D. Poli, et al. Thrombosis Research 183 (2019) 28–32

Table 4
Adverse events during treatment.
VKA DOACs Univariate Multivariatea
(N = 156) (N = 116)

N (x 100 pt-yrs) N (x 100 pt-yrs) HR (95%CI) HR (95%CI)

Bleeding events 6 (1.9) 9 (7.7) 4.7(1.5;15.0) 7.5(2.4;23.6)

Bleeding events at 12 months 1 3 5.1 (0.6;42.1) 6.2 (0.5;71.7)
Major bleedings 2 (0.6) 3 (2.6)
Cerebral 1 0
Gastrointestinal 1c 3
CRNM bleedings 4 (1.3) 6 (5.1)d
Soft tissue ematoma 1 2
Gastrointestinal 2 1
Other 1 3
Thrombotic events 4 (1.3) 4 (3.4) 4.5 (1.5;13.3) 5.7 (1.6;20.3)
Thrombotic events at 12 months 0 4 – –
Recurrent VTE 1 (0.3) 2 (1.7)
Arterial events 3 (1.0) 2 (1.7)b
Death for all causes 50 (16.0) 5 (4.3) 0.3 (0.1;0.9) 0.30(0.1;0.8)
Death at 12 months 15 3 0.33 (0.10;1.13) 0.19 (0.06;0.55)

a
Adjusted for: age at enrollment, Sex, Diabetes mellitus, Hypertension, Frailty; COPD, Previous bleeding, Previous stroke, Active cancer, Renal Failure.
b
Both patients treated with low-dose DOACs.
c
fatal.
d
5 patients on low-dose regimen.

the contrary, as expected, severe renal failure was significantly more the choice between the two anticoagulant regimens. In fact, the more
represented among patients on warfarin. The higher rate of bleeding complex clinical conditions of VKAs patients and the prevalence of
observed in our DOACs cohort could be at least in part due to the higher severe renal failure may be one factor to explain the higher mortality
prevalence of previous bleeding events in this group of patients. rate recorded in these patients.
Similarly, the higher rate of previous cerebral ischemic events in this This study has a number of limitations. Firstly, the observational
group of patients could explain the higher rate of stroke/TIA. Coleman design requires extreme caution in interpreting direct comparisons
et al. [11] reported data of a retrospective claim database analysis of between drugs due the intrinsic limitations and high risks of bias of
patients with VTE with a median age of about 82 years, were the in- these studies. Secondly, frailty was defined by using clinical items for
cidence of major bleeding was similar to that found in our cohort. dementia, bed rest and frequent falls that could be easily reported by
However, in this study no difference was recorded between DOACs and the investigators, without the use of a validated, structured frailty
VKAs. The low number of events recorded in our study could explain stratification score. This explains the relatively low rate of the indicated
this discrepancy, but it should be noted that the median age of our items recorded in our cohort, where only very severe frailties were
patients was markedly higher, reaching about 88 years. Therefore, we recorded. Third, the causes of death reported in the electronic files
cannot exclude that the very high median age of our patients may play a aimed to identify deaths related to bleeding events or cerebral ischemic
role in this observed difference. events. All other causes of death have been defined as not related to
It should be noted that only 5.5% of patients enrolled in our study anticoagulant treatment, including cancer, infectious diseases, vascular
were also on treatment with aspirin; yet the major bleeding rate in this events (not cerebral), heart failure, renal failure, respiratory in-
group was three-fold higher than in patients without aspirin treatment. sufficiency, or sudden death. Finally, there was no central adjudication
Notwithstanding the generally low number of events recorded in the of outcome events in this study.
present study, a low rate of recurrent VTE was recorded during treat-
ment in the two groups of patients, even if higher in patients on DOACs
5. Conclusion
than in those on warfarin.
The use of low-dose regimen has not been tested in the registrative
In conclusion, notwithstanding the long period of treatment widely
trials for the use of DOACs in acute VTE patients, with the exception of
adopted in Italian clinical practice - as it was in our cohort, we found
Hokusai study [12]. We found that about 45% of cases in our popula-
that the rate of bleeding and thrombotic complications among very
tion were receiving a low-dose regimen even in the acute phase of the
elderly VTE patients treated with oral anticoagulants was rather low.
disease. This wide use of low dose regimen did not seem to affect the
Patients treated with DOACs, despite a wide use of low-dose treatment,
efficacy of treatment, considering that only two VTE recurrences were
seemed to have a higher risk for bleeding than patients on warfarin.
recorded among DOACs group, both in patients on full dose regimen.
However, the DOACs low-dose regimens seemed to protect adequately
On the contrary, both the reported stroke/TIA events occurred in low-
the very elderly patients against VTE recurrence. Mortality was mark-
dose DOACs treated patients. In relation to the dose regimens of
edly lower in patients receiving DOACs.
DOACs, we found that five out of nine bleeding events occurred in
patients treated with low doses, suggesting that these patients were not
protected from bleeding risk. Funding
In our cohort, a much lower mortality rate was found among DOACs
treated patients. In RCTs studies comparing DOACs versus warfarin The Arianna Anticoagulation Foundation supported the START2
treatment, no difference in mortality rate was found [13], but it should Register and the study.
be noted that these trials have been conducted among patients with a
median age highly younger than our patients. We may suggest that the
higher mortality rate observed in patients on VKAs could be related, at Declaration of competing interest
least in part, to the selection criteria adopted by physicians in deciding
None declared.

31
D. Poli, et al. Thrombosis Research 183 (2019) 28–32

Appendix A. List of the START2 register investigators Serena Rupoli, Clinica Ematologica, AOU- Ospedali Riuniti Ancona,
Ancona.
Daniela Poli, Rossella Marcucci, Niccolò Maggini, SOD Malattie Giuseppe Malcangi, Centro Emofilia e Trombosi, Policlinico di Bari,
Aterotrombotiche, Azienda Ospedaliero Universitaria-Careggi, Firenze. Bari.
Sophie Testa, Oriana Paoletti, UO Laboratorio Analisi, Centro Maddalena Loredana Zighetti, SIMT- AO San Paolo, Milano.
Emostasi e Trombosi A O Istituti Ospitalieri di Cremona, Cremona. Giovanni Nante, UOS Geriatria ULSS 16 Padova, Padova.
Walter Ageno, Giovanna Colombo, UO. Medicina I, Ospedale di Alberto Tosetto Divisione di Ematologia, Ospedale San Bortolo,
Circolo, Varese. Vicenza.
Benilde Cosmi, Giuliana Guazzaloca, UO di Angiologia e Malattie Rosella Sangiorgio, Centro Emostasi e Trombosi, Ospedale di Lecco,
Coagulazione, AOU S. Orsola-Malpighi, Bologna, Bologna. Lecco.
Anna Falanga, Teresa Lerede, Luca Barcella, USC SIMT, Centro
Emostasi e Trombosi, Ospedale Papa Giovanni XXIII, Bergamo. References
Vittorio Pengo, Gentian Denas, Elisa Bison, Dipartimento di Scienze
Cardio-Toraco-Vascolari, Centro Trombosi, AOU Padova, Padova. [1] S. Barco, Y.W. Cheung, J.W. Eikelboom, M. Coppens, New oral anticoagulants in
Lucia Ruocco, Paolo Chiarugi, Ambulatorio Antitrombosi per la elderly patients, Best Pract Res Clin Haematol. 26 (2013) 215–224.
[2] P. Sardar, S. Chatterjee, S. Chaudhari, G.Y. Lip, New oral anticoagulants in elderly
sorveglianza dei pazienti in terapia anticoagulante orale, U.O. Analisi adults: evidence from a meta-analysis of randomized trials, J Am Geriatr Soc. 62
Chimico-Cliniche, Azienda Ospedaliero Universitaria Pisana, Pisa. (2014) 857–864.
Giuliana Martini, Centro Emostasi, Spedali Civili Di Brescia, Brescia. [3] G. Palareti, D. Poli, The prevention of venous thromboembolism recurrence in the
elderly: a still open issue, Expert Rev Hematol. 11 (2018) 903–909.
Simona Pedrini, Federica Bertola, Servizio di Laboratorio, Istituto [4] E. Antonucci, D. Poli, A. Tosetto, V. Pengo, A. Tripodi, N. Magrini, F. Marongiu,
Ospedaliero Fondazione Poliambulanza Brescia. G. Palareti, The Italian START-Register on anticoagulation with focus on atrial fi-
Lucilla Masciocco, Pasquale Saracino, Angelo Benvenuto, UOC brillation, PLoS One. 10 (2015) e0124719.
[5] C. Kearon, W. Ageno, S.C. Cannegieter, B. Cosmi, G.J. Geersing, P.A. Kyrle,
Medicina Interna, Centro Controllo Coagulazione, Presidio Ospedaliero
Categorization of patients as having provoked or unprovoked venous thromboem-
Lastaria, Lucera (Foggia). bolism: guidance from the SSC of ISTH, J Thromb Haemost. 14 (2016) 1480–1483.
Claudio Vasselli, Laboratorio Patologia Clinica, Policlinico Casilino, [6] F.R. Rosendaal, S.C. Cannegieter, F.J.M. Vandermeer, E. Briet, A method to de-
termine the optimal intensity of oral anticoagulant therapy, Thromb Haemost. 69
Roma.
(1993) 236–237.
Marco Marzolo, UOC Medicina Interna, Ospedale di Rovigo. [7] D.W. Cockcroft, M.H. Gault, Prediction of creatinine clearance from serum creati-
Daniela Mastroiacovo, UOSD Angiologia e Diagnostica Vascolare, nine, Nephron. 16 (1976) 31–41.
Ospedale SS Filippo e Nicola, Avezzano (L'Aquila). [8] R. Pisters, D.A. Lane, R. Nieuwlaat, C.B. de Vos, H.J. Crijns, Lip GY. A novel user-
friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with
Antonietta Piana, Francesco Cibecchini, Centro Prevenzione e Cura atrial fibrillation: the Euro Heart Survey, Chest. 138 (2010) 1093–1100.
Malattie Tromboemboliche Azienda Ospedaliera Universitaria San [9] S. Schulman, C. Kearon, Definition of major bleeding in clinical investigations of
Martino-IST, Genova. antihemostatic medicinal products in non-surgical patients, J Thromb Haemost. 3
(2005) 692–694.
Andrea Toma, Pietro Barbera, UOC di Patologia Clinica, [10] S. Kaatz, D. Ahmad, A.C. Spyropoulos, S. Schulman, Subcommittee on control of A,
Ambulatorio Terapia Anticoagulante Orale, O.C. “L.Cazzavillan” Definition of clinically relevant non-major bleeding in studies of anticoagulants in
Arzignano, (Vicenza). atrial fibrillation and venous thromboembolic disease in non-surgical patients:
communication from the SSC of the ISTH, J Thromb Haemost. 13 (2015)
Antonio Insana, SC Patologia Clinica, Ospedale Santa Croce 2119–2126.
Moncalieri, Moncalieri (Torino). [11] C.I. Coleman, A.G.G. Turpie, T.J. Bunz, J. Beyer-Westendorf, Effectiveness and
Carmelo Paparo, Patologia Clinica, Ospedale Maggiore Chieri safety of rivaroxaban versus warfarin in frail patients with venous thromboembo-
lism, Am J Med. 131 (2018) 933–938 (e1).
(Torino). [12] H.R. Buller, H. Decousus, M.A. Grosso, M. Mercuri, S. Middeldorp, M.H. Prins,
Elvira Grandone, Donatella Colaizzo, Centro Trombosi Casa del G.E. Raskob, S.M. Schellong, L. Schwocho, A. Segers, M. Shi, P. Verhamme,
Sollievo e della Sofferenza, S. Giovanni Rotondo (Foggia). P. Wells, Edoxaban versus warfarin for the treatment of symptomatic venous
thromboembolism, N Engl J Med. 369 (2013) 1406–1415.
Domizio Serra, Centro Trombosi, Ospedale Evangelico
[13] N. van Es, M. Coppens, S. Schulman, S. Middeldorp, H.R. Buller, Direct oral an-
Internazionale, Genova. ticoagulants compared with vitamin K antagonists for acute venous thromboem-
Pasquale Pignatelli, Daniele Pastore, Centro Trombosi, Clinica bolism: evidence from phase 3 trials, Blood. 124 (2014) 1968–1975.
Medica Policlinico Umberto I°, Roma.

32