Introduction
Hyperacute Assessment of Stroke Patients: Time is Brain
Intravenous thrombolysis (IVT) with tissue-type plasminogen activator (recombinant
tissue plasminogen activator [rtPA]) constitutes a state-of-the-art acute, causal
therapy in patients suffering from acute ischemic stroke (AIS). IVT within up to 4.5 hours
of symptom onset in AIS leads to improvement of functional outcome in terms of dependency
after 90 days.[1]
[2]
[3] The use of additional imaging modalities such as perfusion computed tomography (CT)
or magnetic resonance imaging (MRI) can extend the time window of eligibility for
IVT up to 9 hours after onset of stroke or up to 4.5 hours after detection of symptoms
with unknown onset.[4]
[5] However, after stroke onset 1.9 million neurons are destroyed every minute, which
substantiates the paradigm “time is brain” and may explain in parts why delivery of
IVT within 45 minutes of symptom onset is associated with lower 1-year mortality rates.
This was shown in a recent retrospective cohort study in 61,426 patients who received
rtPA within 4.5 hours in which those who were treated within 45 minutes had an all-cause
mortality of 30.8% versus 35.0% respectively (adjusted hazard ratio: 1.13 [95% confidence
interval [CI]: 1.09–1.18]).[6] Consequently, hyperacute stroke care and research has focused on strategies to accelerate
delivery of IVT, which can be quantified by procedural times such as the time from
the arrival of stroke patient in emergency to rtPA administration, also referred to
as “door-to-needle time.”[7] A median door-to-needle time of no more than 30 minutes is associated with a better
functional outcome of AIS patients and is widely considered desirable today.[8]
[9]
[10] Consequentially, strategies to improve speedy delivery of IVT and lower door-to-needle-times
were implemented in stroke care facilities worldwide. These actions include the early
prenotification of appropriate hospitals by emergency medical services in cases of
AIS patients who are potentially eligible for IVT and the immediate availability of
CT or MRI devices for brain imaging. Another strategy is the a priori announcement
of the estimated patient arrival time of an entire stroke team with a single phone
call.[11]
Upon arrival of a potential candidate for IVT, exclusion criteria for IVT must be
identified. Some of these contraindications can be easily ruled out. For instance,
any intracranial bleeding will be detected on cranial CT as part of standard hyperacute
workup and critical arterial hypertension will be detected on routine emergency examination.
However, effective treatment with a direct oral anticoagulant (DOAC), another main
contraindication for IVT, might not be ruled out without a considerable time delay
if a current medication plan is not available and the patient is not capable of communicating
accurate specifications on medication type, dosage, and last time of intake.[12]
[13]
The Role of Pretreatment with Direct Oral Anticoagulants in Decision Making on Intravenous
Thrombolysis for Acute Stroke
DOACs encompass dabigatran (factor IIa inhibitor) as well as rivaroxaban, apixaban,
and edoxaban (factor Xa inhibitors). Direct oral anticoagulation is the standard therapy
for stroke prevention in atrial fibrillation, a condition that goes along with an
up to fivefold increased risk of AIS and makes up 20 to 40% of causes of AIS increasing
with age.[14]
[15] The percentage of patients worldwide who are treated with DOACs is exponentially
increasing.[16]
[17]
[18] This increment also leads to an increase in DOAC-pretreated AIS patients. A recently
published nationwide cohort study conducted in Switzerland reported approximately
20% of 8,179 AIS patients pretreated with DOAC at the time of stroke onset.[19] Furthermore, patients treated with DOAC due to atrial fibrillation are at higher
risk for recurrent AIS compared with patients suffering from atrial fibrillation who
had no prior anticoagulation.[20] Notably, the IVT rate was lower among patients on direct oral anticoagulation in
the Swiss cohort.[19] This may partially be explained by two specific challenges opposed by DOAC pretreatment.
First, there is no global consensus on when IVT in AIS patients on DOAC is considered
safe. Delivery of IVT after at least 48 hours from last DOAC intake is widely accepted
as safe based on the pharmacodynamic and pharmacokinetic profiles of these substances.[21]
[22]
[23] However, this may lead to the exclusion of AIS patients who would still benefit
from IVT without compromising safety of treatment as indicated by a recent meta-analysis
that synthesized data from 52,823 AIS patients.[24] Furthermore, that information is not readily available in the acute setting. It
is noteworthy that recommendations vary between countries. While the American Heart
Association recommends IVT only if screening global coagulation assays or direct factor
Xa activity assays are normal without further quantification,[12] the European Stroke Organization tolerates an uncalibrated anti-Xa activity of the
patient's plasma <0.5 U/mL,[13] which remains highly controversial,[25] or a thrombin time <60 seconds for direct thrombin inhibitors. The French Society
of Vascular Neurology[26] permits any DOAC level < 50 ng/mL if specific test results are available within
30 minutes. A threshold of <30 ng/mL has been adopted for thrombolysis in some expert
opinions.[23]
[27]
[28]
[29]
Second, when DOAC treatment status is unknown, the presence of sub-therapeutic levels
of DOAC needs to be presumed from laboratory results. There are several ways to detect
the presence of DOAC in the patient's blood by routine laboratory assays, and rapid
diagnostic tests for detection of Xa and IIa inhibitors in urine are available as
well. However, the impact of these urine tests is limited for patients potentially
eligible for IVT due to the fact that no quantitative evaluation is possible.[30] Routine blood analyses involve unspecific global coagulation tests and quantitative
measurement for the concentration of specific anticoagulant agent in plasma. Standard
laboratory tests (SLTs) such as activated partial thromboplastin time, prothrombin
time (PT), and thrombin time (TT) are frequently used as global assays to screen for
underlying coagulation factor imbalances and for anticoagulant drug adjustment. SLTs
have some well-known advantages such as standardization, accurate quality control,
and broad availability. However, SLTs show variable sensitivity and specificity for
DOAC depending on the type of DOAC (which may be unknown in the hyperacute evaluation
of stroke patients), because their in vitro effect on SLT does not correlate to the
in vivo anticoagulant activity. Therefore, screening coagulation assays are insufficient
to quantify the degree of DOAC or may fail to detect lower therapeutic DOAC levels.[23]
[31] Quantitative assays include diluted TT (dTT), ecarin clotting time (ECT), ecarin
chromogenic assay (ECA), as well as specific chromogenic anti-IIa and anti-Xa assays.
Dilution of the patient's plasma allows for a more accurate detection of direct IIa
inhibitors in the lower concentration ranges via dTT compared with TT.[23]
[31] Anti-Xa and anti-IIa assays reliably detect therapeutic levels of DOAC in the patient's
plasma and show a linear correlation and agreement with the actual drug concentration.
Diagnostic accuracy of these methods declines at low drug concentrations <30 ng/mL.[23]
[27]
[31] High-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) constitutes
the gold standard for direct measurement of DOAC in the patient's plasma. Since HPLC-MS/MS
is expensive, not routinely available, not suitable for rapid testing, and not easily
operable, laboratory results of dTT and anti-Xa assays using substrate-specific calibrators
and controls serve as the current standard of DOAC detection method in emergency conditions.
Turnaround times (i.e., time between registration of the blood sample in the laboratory
and the communication of the result) for specific anti-Xa testing usually vary between
30 and 90 minutes.[31]
[32]
[33]
[34] Considering a target door-to-needle time of <30 minutes and the narrow time window
for IVT, waiting for laboratory test results imposes a highly relevant delay in IVT
treatment, which compromises functional outcome and, in some cases, even leads to
exceeding the eligibility criteria for IVT.[21]
Point-of-Care Anticoagulation Testing
Prolonged turnaround times motivated the investigation of point-of-care (POC) testing
alternatives in emergency settings. Specific quantitative anti-IIa or anti-Xa assays
do not exist in the POC setting. Examples for POC devices for SLT parameters in within
seconds are CoaguChek POC or the Hemochron Signature Elite POC system, but DOACs cannot
be reliably detected. Sensitivity, specificity, and diagnostic agreement with plasma
drug concentrations measured by the gold standard show high variability depending
on the type of DOAC and results were convincing for edoxaban only.[28]
[35] Viscoelastic testing (VET) of the whole blood is an alternative way to access functional
testing in a POC setting considering all components of coagulation, based on the cellular
model. The (ClotPro) device employs a rotating cup using an elastic element with whole
blood while the pin is stationary. Clotting decelerates the cup rotation due to its
increasing elasticity.[36] In most devices the rotation deceleration is quantified and plotted as a viscoelastometry
amplitude over 40 to 60 minutes. Devices vary in the way they quantify deceleration.
Based on viscoelastometry amplitude characteristics, VET can evaluate clot formation,
firmness, and lysis to estimate different coagulation pathways under approximated
in vivo conditions.
VET devices are frequently used in emergency care medicine where they are predominantly
used in the management of bleeding in trauma or surgical patients.[37]
[38]
[39] Efforts have been made to investigate the capability of thromboelastography and
rotational thromboelastometry devices to identify DOACs. Both devices detect the presence
of oral anticoagulation including vitamin K antagonists, yet discrimination capacity
between different types of oral anticoagulants is not warranted in a nonexperimental
setting and diagnostic accuracy at lower DOAC concentrations in a defined target population
remains to be further investigated.[40]
[41]
[42]
[43] ClotPro (Haemonetics, Boston, Massachusetts, United States) is a recently developed
VET analyzer that is designed to allow drug monitoring and discriminate between different
classes of DOACs by providing additional specific assays (Russell viper venom [RVV]
test and ECA test) that can be run simultaneously with standard assays.[36] A previous exploratory prospective observational study evaluated the ability of
ClotPro to detect DOAC in the patient's plasma at drug concentrations ≥50 and ≥100 ng/mL
in an inhomogeneous population including healthy controls and with low sample size.[44] Another prospective observational study investigated the ability of ClotPro to detect
and quantify DOAC in trauma patients, where timely knowledge about specific DOAC pretreatment
is indispensable for adequate reversal.[45]
Recent efforts have been made to detect DOAC in urine samples using DOAC dipsticks.[46] However, potential correlations with the time of last bladder emptying and the patient's
renal or liver function remain to be investigated. Furthermore, study design limitations
encompass a low sample size of patients with known DOAC status limiting the application
of study results as a screening tool in a real-life scenario where a urine sample
is not readily available as well.[46] Prospectively collected data on the diagnostic accuracy of ClotPro in detecting
DOAC at concentration cut-offs relevant for IVT in a homogeneous study population
with unknown anticoagulation status who are potentially eligible for IVT are missing.
Study Objective
This study aims to determine the diagnostic accuracy of the ClotPro POC device in
detecting whether hyperacute AIS patients have DOAC plasma concentrations that prevent
or allow IVT in a real-life setting based on the consideration of a cut-off at 30 ng/mL
as a clinically relevant concentration of oral anticoagulation in a patient's blood.
Trial Design
This study is a single-center prospective observational diagnostic accuracy study.
Consecutive AIS patients will be screened for the presence of DOAC by quantitative
laboratory assays as well as with HPLC-MS/MS (reference test) and by a ClotPro POC
device with RVV assay or ECA assay (index test) in parallel and once at the time of
presentation at the emergency center, regardless of any documented DOAC treatment
status. In addition, urine samples will be tested with a DOAC dipstick (DOASENSE GmbH,
Heidelberg, Germany) without interfering with the clinical routine.
Study results will be reported in accordance with the STARD 2015 (https://www.equator-network.org/reporting-guidelines/stard/) guidelines.
Methods
Study Setting
This is an investigator-initiated trial that will be conducted at a tertiary stroke
center of the University Hospital Carl Gustav Carus Dresden, Germany. Patients with
AIS who present at the University Emergency Department will be enrolled consecutively.
Eligibility Criteria
Adult patients above 18 years of age who present with imaging or clinically confirmed
AIS will be eligible. Patients with hereditary coagulopathies (von Willebrand disease,
hemophilia A, B, and C, factor VII deficiency, congenital thrombocytopenia) will be
excluded. To avoid prolonged CT due to non-DOAC-related mechanisms of action, we will
also exclude those on therapeutic-dose heparin. Since heparin must be injected intravenously,
subcutaneously, or intranasally, pretreatment with these substances can usually be
ruled out by medical history and does not require POC assessment. Furthermore, trained
technicians of the in-house departments of Anaesthesiology, Forensic Medicine, Clinical
Chemistry, and Laboratory Medicine will perform reference and index tests.
Index Test
ClotPro is a newly developed VET system that uses a cup and a pin to measure clot
formation, with the cup rotating via an elastic element and the pin functioning as
a stationary counterpart.[36] The mechanical deceleration of the cup rotation is detected and translated into
a viscoelastometric amplitude. ClotPro is a bedside-available POC device, mainly used
in the intensive care unit, operation room, or emergency department. Currently there
are nine different test kits available for various measurements depending on the added
reagent, encompassing seven standard assays and two newly developed assays that allow
drug monitoring. All test assays are provided in a ready-to-use application. Dynamics
of clotting can be described by different variables including coagulation time (clotting
time, seconds), clot formation time (seconds), maximum clot firmness (mm), maximum
lysis (%), and lysis time (s). Unlike conventional thromboelastography devices, ClotPro
developed an “active tip technology” where the dry reagent is contained in a sponge
in the tip of the pipette with which the sample is drawn, enabling faster handling
and better reproducibility of test results. Depending on the clinical question, a
particular combination of assays and a particular parameter of clotting need to be
investigated. The recent addition of snake venom RVV-V (RVV factor V activator) or
ECA assays provides a possibility for rapid detection of Xa inhibitors as well as
low-molecular-weight heparin (LMWH; RVV-V) and IIa inhibitors (ECA).[44]
[47] Coagulation time is defined as the time from start of the test until a clot amplitude
of 2 mm is reached. The RVV assay contains a factor X activator derived from the snake
venom “Russell's viper.” The presence of factor Xa inhibitors prolongs the coagulation
time. The reference range for nonaccelerated coagulation time in RVV assay is <80 seconds.
The RVV test does not discriminate between LMWH and factor Xa inhibitors. The ecarin
assay contains ecarin, which activates prothrombin to the thrombin intermediate like
meizothrombin, which is inhibited by direct IIa inhibitors again leading to a prolongation
of the coagulation time. The reference range for coagulation time in the ECA assay
is <100 seconds.[48]
Reference Standard
HPLC-MS/MS directly quantifies DOAC concentrations as well as the concentration of
active metabolites in the patient's plasma. HPLC-MS/MS constitutes a gold standard
in the measurement of DOAC concentrations, yet the high costs, low throughput, and
the need for specifically trained technicians limit its routine use. Hence, the use
of the specified quantitative laboratory assays (dTT and anti-Xa assays) represents
a generally accepted compromise.[23] These assays are comparable to the gold standard when the plasma concentration of
the DOAC is above 30 to 50 ng/mL. However, in lower DOAC levels these tests display
lower accuracy and cannot be used interchangeably with HPLC-MS/MS.[23]
[27]
[31]
[49]
[50]
Blood samples will be collected (one tube per patient) in citrate tubes (S-Monovette,
SARSTEDT AG & Co. KG, Nümbrecht, Germany) and processed at our central laboratory
by trained staff and in accordance with manufacturers' instructions. All analysis
will be performed on STA R Max devices (Stago, Asnières-sur-Seine, France). As part
of clinical routine in every AIS patient with either known negative or unknown DOAC
status dTT will be determined and a LMWH-calibrated anti-Xa assay[51] will be performed as an initial screening tool to identify an anti-Xa activity below
0.35 U/mL, which for every DOAC translates into a concentration <30 ng/mL. In patients
with known DOAC status, specific anti-IIa measurement via dTT or specific anti-Xa
measurements will be run and the quantity of DOAC in the patient's plasma will be
reported as concentration (ng/mL). Additional confirmatory HPLC-MS/MS will be performed
for all samples as reference test to attain exact concentration values.
Anti-IIa Measurement
The concentration of dabigatran will be inferred from the dTT. The dTT resembles the
coagulation time of the patient's diluted plasma. As clotting assay, HEMOCLOT Thrombin
Inhibitor Kit (HYPEN BioMed, CoaChrom Diagnostica GmbH, Maria Enzersdorf, Austria)
is used together with a specific calibrator with dabigatran concentrations of 500,
250, and as low as 50 ng/mL (BIOPHEN Dabigatran Calibrator Low, HYPEN BioMed, CoaChrom
Diagnostica GmbH, Maria Enzersdorf, Austria) and a specific control (BIOPHEN Dabigatran
Control Low, HYPEN BioMed, CoaChrom Diagnostica GmbH, Maria Enzersdorf, Austria) with
known titration for dabigatran. After plotting of the calibration curve with coagulation
time on the y-axis and dabigatran concentration (ng/mL) on the x-axis, the concentration of dabigatran in the patient's plasma can be inferred from
the calibration curve. The lower detection limit is 10 ng/mL.
Anti-Xa Measurement
The concentration of direct anti-Xa inhibitors in the patient's plasma will be inferred
from the one-stage chromogenic assay STA-Liquid Anti-Xa assay (Stago Deutschland GmbH,
Düssedlorf, Germany), which is used together with a drug-specific calibrator and control
(STA-Apixaban Calibrator and STA-Apixaban Control, STA-Rivaroxaban Calibrator and
STA-Rivaroxaban Control, STA-Edoxaban Calibrator and STA-Edoxaban Control, Stago,
Asnières-sur-Seine, France). Chromogenic tests measure the capacity of residual factor
Xa of the patient's plasma to hydrolyze a chromogenic substrate by measuring the change
in optical density per minute, which is inversely correlated with the concentration
of direct Xa inhibitor in the patient's plasma. After plotting a drug-specific calibration
curve with anti-Xa activity (IU/mL) on the y-axis and the concentration of the respective direct Xa inhibitor (ng/mL) on the x-axis, the concentration of the respective direct Xa inhibitor in the patient's plasma
can be inferred from the calibration curve. The lower detection limit is 25 ng/mL
for rivaroxaban, 20 ng/mL for apixaban, and 10 ng/mL for edoxaban.
High-Performance Liquid Chromatography-Tandem Mass Spectrometry
A 500 µL of every blood sample taken in the acute setting will be frozen and multiple
samples will be analyzed in by HPLC-MS/MS using a 1260 Infinity Quaternary LC System
(Agilent Technologies, Santa Clara, California, United States), equipped with a Phenomenex
Luna Pentafluorophenyl Column (length: 150 mm, internal diameter: 2 mm, particle size:
5 µm), and a 3200 Q Trap (AB Sciex Germany GmbH, Darmstadt, Germany) mass spectrometer.[52] Calibration reagents encompass apixaban, edoxaban, and rivaroxaban supplied by Cayman
Chemical Company (Ann Arbor, United States), dabigatran etexilate supplied by Sigma
Aldrich (St. Louis, United States), and dabigatran supplied by Biosynth/Carbosynth
(Thal, Switzerland). Apixaban-13C-d3 (Cayman Chemical Company, Ann Arbor, United States)
is used as the internal standard. The HPLC-MS/MS method for the simultaneous determination
of DOACs that will be used this study has recently been developed and validated by
members of our group.[52]
Outcomes
The primary outcomes of this study are sensitivity and specificity of the ClotPro
POC device in detecting DOAC plasma concentrations above a threshold of 30 ng/mL in
consecutive AIS patients presenting at the emergency department. HPLC-MS will serve
as the reference standard for any DOAC type. ClotPro RVV and ECA assays (index test)
will be run once for every AIS patient by time of presentation and in parallel with
a measurement of the DOAC concentration by the respective reference standard. A dichotomized
test will serve to calculate the diagnostic accuracy of ClotPro to correctly classify
DOAC concentrations in the whole blood as contraindicating or permitting IVT at a
cut-off concentration of 30 ng/mL with respect to the reference standard test. Secondary
study outcomes of diagnostic accuracy comprise positive predictive value (PPV) and
negative predictive value (NPV). Moreover, positive and negative likelihood ratios
will be calculated in terms of proportions. PPV and NPV will be investigated in the
subgroup of patients with assumed DOAC intake and higher pretest probability. In addition,
sensitivity and specificity of the ClotPro device will be investigated for every DOAC
subtype as well as applying a 50 ng/mL threshold as a relevant DOAC plasma concentration.
The correlation and diagnostic agreement between DOAC concentrations as inferred by
standard testing, DOAC dipstick results, and coagulation times of ClotPro assays will
be investigated. Furthermore, turnaround times of reference and index tests as well
as door-to-needle times in case of IVT will be documented as an explorative indicator
for potential time saving. The association of ClotPro data with the risk of bleeding
and clinical outcome in AIS patients will be investigated.
Sample Size
As primary outcome sensitivity and specificity of the ClotPro device to correctly
classify DOAC concentrations in the patient's blood will be investigated, an estimated
overall number of N = 1,850 AIS patients will be consecutively enrolled. Accounting for a drop-out rate
of 15%, a total of n = 1,610 AIS patients need to be analyzed. We performed a sample size calculation
that is sensitive to the prevalence of individual DOACs in our target population of
patients presenting with AIS at the emergency center. The study will therefore be
able to estimate the diagnostic accuracy of POC test results in a real-life setting.[53]
The prevalence of DOAC pretreatment in AIS patients at our local tertiary stroke center
was estimated as follows: in 2020 during an observation period of 200 days, 126 of
580 patients (22%) presenting with AIS were pretreated with oral anticoagulants including
phenprocoumon. Of these, 4.5% were pretreated with edoxaban, 4.0% were pretreated
with rivaroxaban, 6.2% were pretreated with apixaban, and 1.1% were pretreated with
dabigatran. Overall, 15.8% were pretreated with DOAC. These numbers are in line with
external observations on DOAC pretreatment in AIS patients.[17]
[18]
[19]
To determine the diagnostic accuracy of ClotPro test results for the detection of
DOACs in the patient's blood, the clinically acceptable width of a 95% CI is set to
be 10% (w = 0.10). This translates to an acceptance to assume that the true value of the respective
sensitivity or specificity lies with a probability of 95% within a ± 10% range of
the assumed value.
Sensitivity and specificity of ClotPro assays for the detection of anti-Xa inhibitors
determined by plasma-based chromogenic assays have been previously reported in trauma
patients and a mixed group of patients and controls.[44]
[48] In one of these studies,[45] a 50 ng/mL cut-off level for dabigatran plasma concentration was defined clinically
relevant. By contrast, we defined a 30 ng/mL cut-off level relevant to decide whether
IVT can be applied safely in AIS patients. A more conservative guess of the respective
DOAC has been adopted for this study: for edoxaban a sensitivity of 100% and a specificity
of 38% have been reported, for rivaroxaban a sensitivity of 100% and a specificity
of 24% have been reported, for apixaban a sensitivity of 95% and a specificity of
45% have been reported, and for dabigatran a sensitivity and specificity of 100% have
been reported. In case values reached 100%, 98% was used for calculation for direct
anti-Xa inhibitors and 95% for dabigatran.
We proceeded as described in detail by Buderer[53] and received the following results for the number of AIS patients who need to be
consecutively enrolled: for edoxaban n
e = 167, for rivaroxaban n
r = 188, for apixaban n
a = 294, for dabigatran n
d = 960 patients. Because DOAC is mutually exclusive, the total number of consecutive
AIS patients who need to be enrolled adds up to n = 1,610. This number likely represents a slight overestimation of the actual number
needed to analyze because patients taking no DOAC serve as negatives for all DOAC
types. Conversely, the fraction of patients that is positive for one DOAC may not
be accounted for as negative control for a different DOAC.
Recruitment
Patients who are admitted to the emergency department of the University Hospital Carl
Gustav Carus with clinically or imaging confirmed AIS will be consecutively enrolled
in the study. All trial investigators have real-time electronic access to incoming
patient data and will be able to detect AIS candidates by time of presentation. With
an average of 870 patients admitted to our emergency department every year with main
diagnosis of AIS, an overall recruitment period of 26 months is anticipated.
Data Collection
Screening or specific quantitative laboratory assays will be obtained by trained technical
staff at our central laboratory and provided via the clinical information system as
part of the routine diagnostics. Any accessible information on DOAC type and last
time of intake will be made available to the central laboratory to allow for optimal
evaluation of specific dTT and anti-Xa assays as part of clinical routine. ClotPro
assays specific for anti-IIa and anti-Xa activity will run simultaneously to laboratory
assays for all participants. A 500 µL of every sample will be frozen and analyzed
in batches by HPLC-MS/MS. Data will be primarily collected offline in an Excel sheet
(Microsoft Corporation, Albuquerque, New Mexico, United States).
Blinding
Staff assessing the HPLC-MS/MS reference standard will receive number-coded anonymous
probes and thereby be unaware of laboratory as well as ClotPro test results. Results
of the reference standard will not be available to ClotPro assessors due to a delayed
processing of the samples by HPLC-MS/MS. Thereby, a blinding of outcome assessors
will be assured at any time.
Data Management
All clinical data will be transferred to a secure password-protected online database.
Backup copies will be stored offline at the trial center.
Statistical Methods
Descriptive analysis will be used to describe categorical and continuous data. For
continuous data, normal distribution will be checked by means of the Shapiro–Wilk
test. Nonnormally distributed data will be presented using median and interquartile
range and normally distributed continuous data using mean and standard deviation.
Categorical data will be specified using frequency and percentages. Diagnostic accuracy
of the coagulation time in ClotPro RVV and ECA assays in classifying DOAC concentration
levels as obtained by the reference/standard test method at a threshold of 30 ng/mL
in eligible versus noneligible for IVT will be analyzed. Diagnostic accuracy parameters
comprising sensitivity, specificity, PPV, and NPV as well as positive and negative
likelihood ratio for ClotPro test results with respect to the standard test will be
calculated by commonly known means as described elsewhere.[54] Receiver operating characteristic curves will be plotted for coagulation time in
ClotPro RVV and ECA assays and Youden's index will be derived.
Depending on the distribution of anti-Xa and anti-IIa concentrations as well as RVV
and ECA clotting times, Pearson's correlation or Spearman's correlation will be applied
to investigate the correlation between both test results. Regression analysis will
be applied to identify potential risk factors for over- and underestimation by ClotPro
test results with reference to the standard test results. Complete case analysis will
be performed, and missing data will not be imputed. Statistical analysis will be performed
using Stata software (StataCorp 2021, Stata Statistical Software Release 17; StataCorp
LLC, College Station, Texas, United States) and R version 4.0.3 (R Core Team 2021;
R: A language and environment for statistical computing; R Foundation for Statistical
Computing, Vienna, Austria).
Data Monitoring
This is a noninterventional observational study. Diagnostic accuracy of a POC testing
device will be compared with laboratory reference standards. Measurements of the patient's
blood will be performed in vitro, and results will not be incorporated in treatment
decisions. Therefore, data monitoring is not needed.
Harms
Given the nature of the study, no harms are expected.
Auditing
No auditing will be necessary.
