Annals of Internal Medicine
Systematic Review: Benefits and Harms of In-Hospital Use of
Recombinant Factor VIIa for Off-Label Indications
Veronica Yank, MD; C. Vaughan Tuohy, BS; Aaron C. Logan, MD, PhD; Dena M. Bravata, MD, MS; Kristan Staudenmayer, MD, MS;
Robin Eisenhut, BA; Vandana Sundaram, MPH; Donal McMahon, MSc, PhD; Ingram Olkin, PhD; Kathryn M. McDonald, MM;
Douglas K. Owens, MD, MS; and Randall S. Stafford, MD, PhD
Background: Recombinant factor VIIa (rFVIIa), a hemostatic agent
approved for hemophilia, is increasingly used for off-label indications.
Purpose: To evaluate the benefits and harms of rFVIIa use for 5
off-label, in-hospital indications: intracranial hemorrhage, cardiac
surgery, trauma, liver transplantation, and prostatectomy.
Data Sources: Ten databases (including PubMed, EMBASE, and the
Cochrane Library) queried from inception through December 2010.
Articles published in English were analyzed.
Study Selection: Two reviewers independently screened titles and
abstracts to identify clinical use of rFVIIa for the selected indications
and identified all randomized, controlled trials (RCTs) and observa-
tional studies for full-text review.
Data Extraction: Two reviewers independently assessed study char-
acteristics and rated study quality and indication-wide strength of
evidence.
Data Synthesis: 16 RCTs, 26 comparative observational studies,
and 22 noncomparative observational studies met inclusion criteria.
Identified comparators were limited to placebo (RCTs) or usual care
(observational studies). For intracranial hemorrhage, mortality was
R ecombinant factor VIIa (rFVIIa) is an expensive and
potent procoagulant. The U.S. Food and Drug Ad-
ministration (FDA) approved intravenous use of rFVIIa in
1999 for patients with hemophilia A or B and antibody in-
hibitors against standard-factor replacements. Recently, its use
has expanded beyond these approved indications to encom-
pass a wide range of in-hospital, off-label applications.
Off-label drug use refers to any application that devi-
ates from FDA-approved use. The FDA drug-approval
process mandates that randomized, clinical trials demon-
strate efficacy and safety. Once approval is given, however,
physicians are free to use the drug for other indications.
Although off-label use is legal and allows for rapid adop-
tion of some therapies, the available evidence supporting it
usually falls short of the rigor that accompanies FDA re-
view. Even though the resulting uncertainty may be accept-
able, concerns increase when off-label use is applied to
conditions that are clinically distinct from approved indi-
cations or it is done frequently, is costly, or is associated
with important side effects or harms.
The off-label use of rFVIIa for hospitalized patients
has increased, despite concerns about efficacy and safety,
including evidence suggesting an increased rate of throm-
boembolic events (1–5). Our comparative effectiveness re-
view evaluates the benefits and harms of in-hospital, off-
label use of rFVIIa in adults for the selected indications of
www.annals.org
Review
not improved with rFVIIa use across a range of doses. Arterial
thromboembolism was increased with medium-dose rFVIIa use (risk
difference [RD], 0.03 [95% CI, 0.01 to 0.06]) and high-dose rFVIIa
use (RD, 0.06 [CI, 0.01 to 0.11]). For adult cardiac surgery, there
was no mortality difference, but there was an increased risk for
thromboembolism (RD, 0.05 [CI, 0.01 to 0.10]) with rFVIIa. For
body trauma, there were no differences in mortality or thrombo-
embolism, but there was a reduced risk for the acute respiratory
distress syndrome (RD, ⫺0.05 [CI, ⫺0.02 to ⫺0.08]). Mortality was
higher in observational studies than in RCTs.
Limitations: The amount and strength of evidence were low for
most outcomes and indications. Publication bias could not be
excluded.
Conclusion: Limited available evidence for 5 off-label indications
suggests no mortality reduction with rFVIIa use. For some indica-
tions, it increases thromboembolism.
Primary Funding Source: Agency for Healthcare Research and
Quality.
Ann Intern Med. 2011;154:529-540. www.annals.org
For author affiliations, see end of text.
intracranial hemorrhage (ICH), cardiac surgery, trauma,
liver transplantation, and prostatectomy.
METHODS
The Agency for Healthcare Research and Quality
(AHRQ) commissioned the full report, which is available,
including the search strategies and detailed evidence tables,
at the AHRQ Web site (6). We developed and followed
standardized protocols for data searches, extraction, quality
assessments, and syntheses.
See also:
Print
Editorial comment. . . . . . . . . . . . . . . . . . . . . . . . . . 566
Related article. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516
Web-Only
Appendix Tables
Appendix Figure
CME quiz
Conversion of graphics into slides
19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 529
Review Effectiveness of Off-Label Use of Recombinant Factor VIIa
Data Sources and Searches
Searches
In collaboration with a research librarian, we devel-
oped individualized search strategies for 10 biblio-
graphic databases from inception through 31 December
2010: PubMed, EMBASE, the Cochrane Library, ACP
Journal Club, Database of Abstracts of Reviews of Ef-
fects, Cochrane Central Register of Controlled Trials,
CMR, Health Technology Assessment, National Health
Service Economic Evaluation Database, and BIOSIS.
We contacted experts and reviewed bibliographies of
identified systematic reviews, files supplied by the man-
ufacturer, and the manufacturer’s Web site. A librarian
expert on gray literature (sources other than published
materials indexed in bibliographic databases) searched
regulatory sites, clinical trial registries, conference pro-
ceedings, and grant-funded and federally funded re-
search sites and contacted authors of abstracts to deter-
mine whether full reports had been subsequently
published.
Inclusion Criteria
We sought studies that compared the use of rFVIIa
with alternative therapies, placebo, or usual care for hospi-
talized patients with 5 off-label indications: ICH, cardiac
surgery, trauma, liver transplantation, and prostatectomy.
For inclusion, studies had to address direct or surrogate
clinical outcomes. We did not contact authors about re-
sults that were not reported in their publications. We ex-
cluded studies published only as abstracts and studies of
human factor VIIa or modified recombinant forms under
development; studies on use of rFVIIa for on-label indica-
tions and rFVIIa applied to patients substantially similar to
those for whom on-label indications are approved (for ex-
ample, Glanzmann thrombasthenia); and studies whose
outcome measures (for example, drug half-life) are not rel-
evant to efficacy, effectiveness, or safety. Lastly, we re-
viewed the English-language abstracts and references of
studies published in non-English languages. Among these,
we identified 2 small comparative studies likely eligible for
effectiveness review (randomized, controlled trial [RCT] of
cardiac surgery [11 patients who received rFVIIa] [7] and
comparative cohort study of brain trauma [7 who received
rFVIIa] [8]) and 6 case series reports (range, 16 to 35
patients) (9 –14) likely eligible for harms analyses if we had
the capacity to do a full-text review. The RCT was used
for sensitivity testing, with similar results to those re-
ported here. The other articles not published in English
were excluded from further review on the basis of a lack
of capacity for translation and the determination that
such exclusions were unlikely to bias our findings be-
cause few articles were identified and the studies en-
rolled few patients.
530 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8
Study Selection
Two reviewers independently screened titles and ab-
stracts to identify clinical use of rFVIIa for the selected
indications and identified all RCTs and observational stud-
ies for full-text review. For the effectiveness analyses, we
included RCTs and comparative observational studies. We
reviewed all RCTs in detail, whereas we reviewed only
comparative observational studies of fair or good quality in
detail. We used comparative observational studies of poor
quality for qualitative sensitivity analyses. Patients from
overlapping or duplicate articles reporting on the same pa-
tient population were included only once.
Although noncomparative studies are a less reliable
source of evidence than comparative studies, they may re-
port infrequent events not identified in RCTs (15). Given
concerns about the possibility of clinically significant
harms, we included registries and cohorts with at least 15
total patients in the harms analyses. We selected 15 pa-
tients as the minimum threshold because the risk for bias
(for example, selective reporting) is probably increased in
smaller reports. We also included the treatment groups of
all comparative studies in the harms analyses.
Data Extraction and Quality Assessment
Two reviewers independently rated study quality and
indication-wide strength of evidence and abstracted study
characteristics by using pretested electronic forms. Study
quality and indication-wide strength of evidence were eval-
uated by using well-established criteria (16 –25) and a pre-
defined systematic approach described further in the
AHRQ report (6). The quality criteria were specific to the
study type and were used to assign each study a quality
score from the ordinal scale of poor, fair, or good by using
qualitative determinations rather than numerical scores.
For RCTs, essential quality elements included appropriate
randomization, allocation concealment, blinding, and ab-
sence of differential follow-up. For observational studies,
essential quality elements included appropriate methods to
generate comparability of groups and control for con-
founding, appropriate blinding, and absence of differential
follow-up. Studies that did not achieve (or did not ade-
quately report) these elements were assigned a quality score
of poor. Studies that achieved these partially or fully were
felt to have a lower risk for bias and were assigned quality
scores of fair or good, respectively. Strength of evidence
(low, moderate, high, or insufficient) was determined sim-
ilarly (17). Five types of data were abstracted from each
included study: study design, population evaluated, rFVIIa
dosing and administration, outcomes assessed, and study
funding. Disagreements were resolved by discussion and,
when necessary, review by a third author.
Data Synthesis and Analysis
We summarized mortality, thromboembolism, and
other outcome rates for included studies. We considered
studies eligible for meta-analysis if they were good- or fair-
quality RCTs or good-quality observational studies and
www.annals.org
 Effectiveness of Off-Label Use of Recombinant Factor VIIa
had similar interventions and patient populations. We did
meta-analyses when there were at least 2 studies of fair or
better quality, including at least 1 of good quality.
The ICH RCTs had several intervention groups in
which varying doses of rFVIIa were compared with a single
control group. For these, we used a least-square fixed-
effects model, a standard meta-analytic methodology for
synthesizing studies of this design (see the Gleser–Olkin
model [26 –28]). We did meta-analyses by using both
fixed-effects and random-effects models for sensitivity test-
ing. We calculated standardized mean differences for per-
centage of hematoma expansion, the only continuous vari-
able analyzed. For dichotomous outcomes, we calculated 2
effect-size metrics: risk differences (RDs) and arcsine stan-
dardized mean differences (29). Results from these metrics
were consistent. For ease of interpretation and because out-
come rates were similar across studies (such that disadvan-
tages of the RD metric were minimized), only RDs are
reported here. We did meta-analytic calculations with the
R statistics package, version 2.11.1 (www.r-project.org), by
using a modified meta package (Guido Schwarzer, sc@imbi
.uni-freiburg.de, version 1.6-0). Although we did assess-
ments of heterogeneity by using the Q and I2 statistics, we
expected these to be nonsignificant if too few studies were
included in each meta-analytic calculation. In that case,
differences in findings between the fixed- and random-
effects analyses were expected to highlight the presence of
heterogeneity.
Because the literature suggested a dose–response rela-
tionship between rFVIIa and certain outcomes, particularly
arterial thromboembolism, and the ICH RCTs reported
outcomes separately for several rFVIIa doses and arterial
versus venous thromboemboli, we chose a priori to analyze
these data according to low, medium, and high doses (ⱕ40
mcg/kg, ⬎40 but ⬍120 mcg/kg, and ⱖ120 mcg/kg) and
arterial versus venous thromboembolic outcomes. Dose–
response analyses were not possible for other indications
because there were too few patients at specific dosages,
dosages were unclear, or outcomes were not reported by
dosage.
Role of the Funding Source
Primary funding for the project was provided by the
AHRQ, with additional support from the National Heart,
Lung, and Blood Institute and the Palo Alto Medical
Foundation Research Institute. Although the AHRQ formu-
lated the initial study questions, the funding sources otherwise
had no role in the design and conduct of this study or in the
collection, management, analysis, or interpretation of the
data. The AHRQ reviewed the manuscript but did not assist
in its preparation. The funding sources had no role in the
decision to submit the manuscript for publication.
RESULTS
Our searches identified 6191 potentially relevant arti-
cles, of which 62 articles reporting on 64 studies met in-
www.annals.org
Review
clusion criteria (Appendix Figure, available at www.annals
.org). Of these, 26 studies (16 RCTs [30 – 43], 10
comparative observational studies [44 –53]) are included in
the comparative effectiveness review (Table). Most had
small-to-moderate sample sizes (for patients who received
rFVIIa, mean, 80; median, 44; range, 6 to 573). There was
great variability in doses of rFVIIa administered (range, 5
to 400 mcg/kg). Thirty-eight additional studies (16 poor-
quality comparative [54 – 69] and 22 noncomparative [70 –
91] observational studies) were included in the harms anal-
yses. Overall, no studies compared rFVIIa with predecessor
products that might be considered alternatives for some
indications (for example, activated prothrombin complex
concentrates). Instead, the RCTs examined rFVIIa versus
placebo and the comparative observational studies com-
pared it with usual care. There were few fair- or good-
quality studies within any indication (Table) (6). Random-
ized, controlled trials were generally of better quality than
comparative observational studies: Only 13% (2 of 16) of
RCTs (41, 42) were of poor quality, whereas 62% (16 of
26) of the comparative observational studies (54 – 69) were
of poor quality. All observational studies that were deter-
mined to be of poor quality lacked appropriate methods to
generate comparability of groups or control for confound-
ing. The small number of studies for any given indication
precluded the use of funnel plots or other approaches for
evaluation of publication bias. The manufacturer of
rFVIIa, Novo Nordisk (Bagsværd, Denmark), sponsored
most of the RCTs.
Effectiveness Review Outcomes
The Table summarizes the key characteristics of the
studies providing comparative effectiveness outcome
data, along with their quality, strength of evidence, and
conclusions. Although Figure 1 summarizes the mortal-
ity and thromboembolic event risk differences in the
individual studies, few of these studies were powered to
distinguish important differences between treatment
groups for these direct end points. Instead, they used
indirect end points (for example, transfusion require-
ments) as primary outcomes. For all indications, quali-
tative sensitivity analyses suggest that poor-quality com-
parative observational studies (those not reviewed in
depth in the effectiveness review but included in the
harms analysis [Appendix Tables 1 to 12, available at
www.annals.org]) had results consistent with those re-
ported for each indication for the RCTs and higher-
quality observational studies (6). In the sections that
follow, we present the benefits and harms associated
with use of rVIIa by specific indication. In all cases in
which meta-analyses were done, the results of the fixed-
effects and random-effects models were virtually identi-
cal and were in agreement on all determinations of sig-
nificant versus nonsignificant findings (that is, those
that crossed the null), indicating the presence of little
heterogeneity among included studies. We report results
19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 531
Review Effectiveness of Off-Label Use of Recombinant Factor VIIa

Table. In-Depth Assessment of Studies in Effectiveness Review: Outcomes, Strength of Evidence, and Conclusions*

Study, Year (Reference) Total Total Total Patients, Outcome Strength of Range of Effect Sizes Conclusions (Association
RCTs or Studies, by by Group, n Evidence for Risk per Period, by of Use of rFVIIa vs. Usual
OBS, n Quality, n Group† Care)

rFVIIa Usual rFVIIa Usual

Group Care Group Care
Group Group
Intracranial hemorrhage
Mayer et al, 2005 (30) and RCT: 4 Good: 2 968 384 Mortality Moderate 0.08–0.22 0.13–0.29 No effect on mortality
2008 (33) (Figure 3)
Mayer et al, 2005 (31) and Fair: 2 Arterial TE events Moderate 0.04–0.11 0–0.13 Increased rate of arterial
2006 (32) TE events (Figure 3)
Ilyas et al, 2008 (44) OBS: 1 Fair: 1 Poor functional Moderate 0.44–0.53 0.46–0.69 No effect on poor
status‡ functional status
Percentage of Moderate 4–79 11–29 Decrease in percentage of
hematoma hematoma expansion
expansion

Cardiac surgery
Gill et al, 2009 (35) RCT: 2 Good: 1 251 216 Mortality Low 0–0.33 0.06–0.33 No effect on mortality
(Figure 3)
Diprose et al, 2005 (34) Fair: 1 TE events Moderate 0–0.22 0–0.20 Increased rate of TE events
(Figure 3)
Karkouti et al, 2005 (45), and OBS: 4 Good: 2 RBCs transfused Low 0–9.1 2–17 Possible reduction in RBC
Gelsomino et al, 2008 (46) (units) transfusion requirements
von Heymann et al, 2005 (47), Fair: 2 ICU length of stay Low 2.5–14 1–18.5 Unclear effect on ICU
and Tritapepe et al, 2007 (48) (d) length of stay

Body trauma
Hauser et al, 2010 (37) RCT: 4 Good: 2 746 900 Mortality Moderate 0.07–0.31 0–0.51 No effect on mortality
(Figure 3)
Boffard et al, 2005 (36) Fair 2 TE events Moderate 0.03–0.20 0–0.14 No effect on TE event rate
(Figure 3)
Rizoli et al, 2006 (49); Fox et al, OBS: 3 Fair: 3 ARDS Moderate 0–0.06 0.04–0.16 Decrease in rate of ARDS
2009 (50); and Wade et al, (Figure 3)
2010 (51)
RBCs transfused Moderate 5.0–16.0 6.8–14.0 Unclear effect on RBC
(units) transfusion requirements

Brain trauma
Narayan et al, 2008 (38) RCT: 1 Fair: 1 79 53 Mortality Low 0.12–0.33 0.11–0.53 No effect on mortality
TE events Low 0.15–0.22 0.08–0.18 No effect on TE event rate
Stein et al, 2008 (52) OBS: 1 Fair: 1 Absolute hematoma Low 7.0 10.4 No decrease in hematoma
expansion (mL) growth

Liver transplantation
Lodge et al, 2005 (39), and RCT: 4 Fair: 2 215 117 Mortality Low 0–0.08 0–0.02 No effect on mortality
Planinsic et al, 2005 (40)
Pugliese et al, 2007 (41), and Poor: 2 TE events Low 0–0.22 0–0.16 No effect on TE event rate
Liu et al, 2009 (42)
Hendriks et al, 2001 (53) OBS: 1 Fair: 1 RBCs transfused Low 1.2–13.0 2.3–11.1 Possible reduction in RBC
(units) transfusion requirements
OR time (min) Low 268–554 432–598 No reduction in OR time
ICU length of stay Low 3.0–4.8 3.0–5.2 No reduction in ICU
(d) length of stay

Prostatectomy
Friederich et al, 2003 (43) RCT: 1 Fair: 1 24 12 Mortality Insufficient 0 0 Cannot comment because
of limited events
TE events Insufficient 0–0.13 0 Cannot comment because
of limited events
RBCs transfused Insufficient 0–0.6 1.5 Reduction in RBC
(units) transfusion requirements
OR time (min) Insufficient 120–126 180 Reduction in OR time

ARDS ⫽ acute respiratory distress syndrome; ICU ⫽ intensive care unit; OBS ⫽ comparative observational study; OR ⫽ operating room; RBC ⫽ red blood cell; RCT ⫽
randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa; TE ⫽ thromboembolic.
* The effectiveness review evaluates in depth all RCTs (30 – 43) and fair- or better-quality comparative observational studies (44 –53). Strength of evidence is based on scores
within 4 evidence domains (risk for bias, consistency, directness, and precision) and is rated as low, moderate, high, or insufficient.
† Outcome is given as a range of rates, unless otherwise stated. Each outcome range encompasses the lowest and highest rate per unit measured across all studies and, as such,
should not be used to directly compare between the rFVIIa and usual care groups. Direct comparisons between groups are described in detail in the main report and are
summarized in the conclusions column.
‡ Poor functional status is a modified Rankin scale score of 4 to 6.

532 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 www.annals.org
 Effectiveness of Off-Label Use of Recombinant Factor VIIa Review

from the fixed-effects models. Unless otherwise noted,

Figure 1. Mortality and thomboembolic event risk
the low strength of evidence critically limits our ability
differences (rFVIIa minus usual care) for indications with at
to draw conclusions about the comparative effectiveness
least 2 comparative studies included in the effectiveness
of rFVIIa and usual care.
review.

0.15
ICH
Identified RCTs examined only intracerebral hemor- 0.10
rhage, rather than other forms of ICH (for example, sub-
dural hematoma). Four RCTs (30 –33) and 1 comparative 0.05

Mortality Risk Difference

observational study (44) examined the use of rFVIIa in 968 0.00
intervention patients (Table). Notably, although the RCTs
excluded patients receiving oral anticoagulation, all of the –0.05
patients in the 1 comparative observational study that met
–0.10
inclusion criteria were receiving oral anticoagulants— half
of whom had intracerebral hemorrhage and the other half –0.15
had subdural hematoma. The meta-analyses found no evi-
dence of effect on mortality of rFVIIa at any dosing level –0.20

(Figure 2 and Appendix Tables 1 to 3) or poor functional –0.25

outcome (as measured by the modified Rankin scale) (RD
for low dose, ⫺0.02 [95% CI, ⫺0.09 to 0.05]; medium 0.20
dose, ⫺0.03 [CI, ⫺0.10 to 0.04]; and high dose, ⫺0.04
[CI, ⫺0.15 to 0.08]) compared with usual care. However, 0.15
there was an increased rate of arterial thromboembolism
with rFVIIa for the medium- and high-dose groups (Figure
TE Event Risk Difference

0.10

3). Use of rFVIIa significantly decreased relative hematoma

expansion at all doses (standardized mean difference for 0.05

low dose, ⫺0.15 [CI, ⫺0.29 to 0.00]; medium dose,

⫺0.24 [CI, ⫺0.39 to ⫺0.10]; and high dose, ⫺0.33 [CI, 0.00

⫺0.58 to ⫺0.09]). On statistical testing, there was no ev- –0.05

idence of dose effect for any end point, including mortality
and arterial thromboembolism (6). Thus, moderate
strength of evidence suggests that the use of rFVIIa for
patients with ICH who were not receiving oral anticoagu-
lation yields no significant benefit for mortality or func-
tional outcome but is associated with an increased risk for
arterial thromboembolism.
Cardiac Surgery
Two RCTs (34, 35) and 4 comparative observational
studies (45– 48) of 251 adult cardiac surgery patients re-
ceiving rFVIIa met inclusion criteria, of which the 2 RCTs
and 2 of the observational studies (45, 46) were of suffi-
cient quality for meta-analysis. One RCT assessed prophy-
lactic rFVIIa use at the conclusion of complex, noncoro-
nary artery bypass grafting surgeries (34); the other studies
evaluated treatment for postoperative bleeding. However,
the timing and context of rFVIIa administration was sim-
ilar for all studies because bleeding is common after com-
plex, noncoronary artery bypass grafting surgery, and
rFVIIa was given after cardiopulmonary bypass in all cases.
Thus, we considered it appropriate to combine these stud-
ies. In meta-analyses, we found no effect of rFVIIa on
mortality compared with usual care but an increased risk
for thromboembolism (Figures 2 and 3 and Appendix Ta-
www.annals.org
–0.10
–0.15
ICH Cardiac Body Brain Liver
Surgery Trauma Trauma Transplantation
Indications with 2 or more comparative studies are included in the
figure: ICH (30 –33, 44), cardiac surgery (34, 35, 45– 48), body
trauma (36, 37, 49 –51), brain trauma (38, 52), and liver transplan-
tation (39 – 42, 53). Each circle represents a study. Larger circles
correspond to larger studies, shaded circles represent studies on treat-
ment use of rFVIIa, and open circles represent studies on prophylac-
tic use of rFVIIa. Intracranial hemorrhage outcomes here reflect total
TE events in contrast to the arterial TE events assessed in the meta-
analyses. For cardiac surgery, 3 study circles overlap at the 0 abscissa
for mortality risk, and 2 similarly overlap for TE event risk. ICH ⫽
intracranial hemorrhage; rFVIIa ⫽ recombinant factor VIIa; TE ⫽
thromboembolic.
ble 4). Red blood cell transfusion requirements were pos-
sibly reduced with rFVIIa, but the trend was apparent only
across higher-quality studies (Appendix Table 5). Results
about length of stay in the intensive care unit were not
consistent. In summary, current evidence of moderate
strength (for thromboembolic events) or low strength (for
all other outcomes) suggests no mortality benefit, but that
rFVIIa use increases the risk for thromboembolism.
19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 533
Review Effectiveness of Off-Label Use of Recombinant Factor VIIa

Figure 2. Meta-analysis of mortality associated with off-label use of rFVIIa for ICH, adult cardiac surgery, and body trauma in
RCTs and good-quality observational studies.

Study, Year (Reference) rFVIIa Usual Care RD (95% CI)

Events/Total, n/n Events/Total, n/n
ICH, low dose
Mayer et al, 2005 (30) 19/108 28/96 –0.12 (–0.23 to 0.00)
Mayer et al, 2005 (31) 1/18 2/11 –0.13 (–0.38 to 0.13)
Mayer et al, 2006 (32) 7/24 1/8 0.17 (–0.13 to 0.46)
Mayer et al, 2008 (33) 50/276 51/268 –0.01 (–0.07 to 0.06)
Combined 426 383 –0.03 (–0.09 to 0.02)
ICH, medium dose
Mayer et al, 2005 (30) 17/92 28/96 –0.11 (–0.23 to 0.01)
Mayer et al, 2005 (31) 0/6 2/11 –0.18 (–0.48 to 0.12)
Mayer et al, 2006 (32) 0/8 1/8 –0.12 (–0.41 to 0.16)
Mayer et al, 2008 (33) 62/297 51/268 0.02 (–0.05 to 0.08)
Combined 403 383 –0.02 (–0.08 to 0.04)
ICH, high dose
Mayer et al, 2005 (30) 20/103 28/96 –0.10 (–0.22 to 0.02)
Mayer et al, 2005 (31) 2/12 2/11 –0.02 (–0.33 to 0.30)
Combined 115 107 –0.02 (–0.11 to 0.07)
Cardiac surgery
Diprose et al, 2005 (34) 0/9 1/10 –0.10 (–0.34 to 0.14)
Gill et al, 2009 (35) 10/103 4/69 0.04 (–0.04 to 0.12)
Karkouti et al, 2005 (45) 7/51 7/51 0.00 (–0.13 to 0.13)
Gelsomino et al, 2008 (46) 2/40 3/40 –0.02 (–0.13 to 0.08)
Combined 203 170 0.01 (–0.05 to 0.06)
Body trauma
Boffard et al, 2005 (36) (blunt) 30/69 35/74 –0.04 (–0.20 to 0.12)
Hauser et al, 2010 (37) (blunt) 24/218 26/242 0.00 (–0.05 to 0.06)
Boffard et al, 2005 (36) (penetrating) 29/70 28/64 –0.02 (–0.19 to 0.14)
Hauser et al, 2010 (37) (penetrating) 8/44 5/38 0.05 (–0.11 to 0.21)
Combined 401 418 0.00 (–0.05 to 0.05)

–0.4 –0.2 0 0.2 0.4

RD (95% CI)
Favors rFVIIa Favors Usual Care

For ICH, all studies are RCTs (30 –33) and meta-analyses are done according to dosing category (low, medium, and high). For cardiac surgery, the
studies by Diprose and colleagues (34) and Gill and coworkers (35) are RCTs, whereas those by Karkouti and colleagues (45) and Gelsomino and
coworkers (46) are observational studies. For body trauma, all studies are RCTs (36, 37). ICH ⫽ intracranial hemorrhage; RCT ⫽ randomized,
controlled trial; RD ⫽ risk difference; rFVIIa ⫽ recombinant factor VIIa.

Body Trauma not report usable comparative data on this outcome.

Four RCTs (published in 2 articles that both report
on 1 blunt and 1 penetrating trauma trial) (36, 37) and
3 comparative observational studies (49 –51) assessed
746 patients who received rFVIIa. Two RCTs (36) cen-
sored patients who died within 48 hours of injury for
analyses of transfusion requirements—a critical limita-
tion of this literature— but report on the end points of
mortality, thromboembolic events, and the acute respi-
ratory distress syndrome (ARDS) in all patients. In
meta-analyses of the 4 RCTs, we found no effect of
rFVIIa on mortality or thromboembolism but a signifi-
cant reduction in ARDS (Figures 2 and 3 and Appendix
Table 6). Use of rFVIIa reduced transfusion require-
ments in all 4 RCTs when all patients were included in
the analysis (that is, including previously censored pa-
tients), but the decrease was significant in only 1 RCT
(37) (Appendix Table 7). The observational studies did
534 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8
Overall, available evidence of moderate strength identi-
fies no increased risk for thromboembolic events, a re-
duced risk for ARDS, and no difference in mortality.
Brain Trauma
One RCT (38) and 1 comparative observational study
(52) evaluated 79 patients who had traumatic brain injury
and received rFVIIa. The observational study included pa-
tients receiving oral anticoagulation, whereas the RCT ex-
cluded them and also patients for whom neurosurgery
was planned. On systematic review, we found no effect
of rFVIIa on mortality or thromboembolic events (RD
[rFVIIa minus usual care] for mortality, 0 in the RCT and
⫺0.2 in the observational study; RD for thromboembolic
events, 0.07 and 0.05, respectively) (Figure 1 and Appen-
dix Table 8). In addition, rFVIIa use did not influence
www.annals.org
 Effectiveness of Off-Label Use of Recombinant Factor VIIa Review

hematoma growth but was associated with a reduction in
time to neurosurgical intervention (that is, by more quickly
normalizing the international normalized ratio) in the ob-
servational study, the 1 study that evaluated this outcome.
In summary, current evidence of low strength is too lim-
ited to compare the harms and benefits of rFVIIa.
Liver Transplantation
Four RCTs (39 – 42) and 1 comparative observational
study (53) evaluated 215 patients with Child class B or C
cirrhosis who received prophylactic rFVIIa on initiation of
liver transplantation. On systematic review, we found no
effect of rFVIIa on mortality or thromboembolism (RD
range [rFVIIa minus usual care] for mortality, 0 to 0.01;
RD range for thromboembolic events, ⫺0.03 to 0.17)
(Figure 1 and Appendix Table 9). There was a trend across
studies toward reduced red blood cell transfusion require-
ments with prophylaxis, but neither operating room time
nor length of stay in the intensive care unit was reduced
(Appendix Table 10). Thus, available evidence of low

Figure 3. Meta-analysis of thromboembolic events and the acute respiratory distress syndrome associated with off-label use of
rFVIIa for some indications in RCTs and good-quality observational studies.

Study, Year (Reference) rFVIIa Usual Care RD (95% CI)

Thromboembolic events Events/Total, n/n Events/Total, n/n
ICH, low dose
Mayer et al, 2005 (30) 6/108 0/96 0.06 (0.01 to 0.10)
Mayer et al, 2005 (31) 3/18 0/11 0.17 (–0.04 to 0.37)
Mayer et al, 2006 (32) 0/24 1/8 –0.12 (–0.37 to 0.12)
Mayer et al, 2008 (33) 15/265 12/263 0.01 (–0.03 to 0.05)
Combined 415 378 0.02 (0.00 to 0.05)
ICH, medium dose
Mayer et al, 2005 (30) 2/92 0/96 0.02 (–0.01 to 0.06)
Mayer et al, 2005 (31) 0/6 0/11 0.00 (–0.22 to 0.22)
Mayer et al, 2006 (32) 0/8 1/8 –0.12 (–0.41 to 0.16)
Mayer et al, 2008 (33) 27/293 12/263 0.05 (0.00 to 0.09)
Combined 399 378 0.03 (0.01 to 0.06)
ICH, high dose
Mayer et al, 2005 (30) 8/103 0/96 0.08 (0.02 to 0.13)
Mayer et al, 2005 (31) 0/12 0/11 0.00 (–0.15 to 0.15)
Combined 115 107 0.06 (0.01 to 0.11)
Cardiac surgery
Diprose et al, 2005 (34) 2/9 2/10 0.02 (–0.35 to 0.39)
Gill et al, 2009 (35) 7/103 1/69 0.05 (0.00 to 0.11)
Karkouti et al, 2005 (45) 8/51 5/51 0.06 (–0.07 to 0.19)
Gelsomino et al, 2008 (46) 2/40 0/40 0.05 (–0.03 to 0.13)
Combined 203 170 0.05 (0.01 to 0.10)
Body trauma
Boffard et al, 2005 (36) (blunt) 2/69 3/74 –0.01 (–0.07 to 0.05)
Hauser et al, 2010 (37) (blunt) 45/218 35/242 0.06 (–0.01 to 0.13)
Boffard et al, 2005 (36) (penetrating) 4/70 3/64 0.01 (–0.06 to 0.09)
Hauser et al, 2010 (37) (penetrating) 2/44 5/38 –0.09 (–0.21 to 0.04)
Combined 401 418 0.01 (–0.03 to 0.04)

The acute respiratory distress syndrome

Body trauma
Boffard et al, 2005 (36) (blunt) 3/69 12/74 –0.12 (–0.22 to –0.02)
Hauser et al, 2010 (37) (blunt) 8/218 18/242 –0.04 (–0.08 to 0.00)
Boffard et al, 2005 (36) (penetrating) 4/70 5/64 –0.02 (–0.11 to 0.06)
Hauser et al, 2010 (37) (penetrating) 0/44 3/38 –0.08 (–0.17 to 0.02)
Combined 401 418 –0.05 (–0.08 to –0.02)

–0.4 –0.2 0 0.2 0.4

RD (95% CI)
Favors rFVIIa Favors Usual Care

For ICH, all studies are RCTs (30 –33), meta-analyses are done according to dosing category (low, medium, and high), and analyses of thromboembolic
events are for arterial events only. For cardiac surgery, the studies by Diprose and colleagues (34) and Gill and coworkers (35) are RCTs, whereas those
by Karkouti and colleagues (45) and Gelsomino and coworkers (46) are observational studies, and the meta-analyses of thromboembolic events evaluate
all events (both arterial and venous). For body trauma, all studies are RCTs (36, 37), and the meta-analyses of thromboembolic events evaluate all events.
ICH ⫽ intracranial hemorrhage; RCT ⫽ randomized, controlled trial; RD ⫽ risk difference; rFVIIa ⫽ recombinant factor VIIa.
www.annals.org 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 535
Review Effectiveness of Off-Label Use of Recombinant Factor VIIa

Figure 4. Harms analysis: weighted mortality and thromboembolic events in all RCTs and observational studies, by indication.

Liver
ICH Cardiac Surgery Body Trauma Brain Trauma Transplantation Prostatectomy
Weighted Mean Mortality Rate

0.50

0.40

0.30

0.20

0.10

0.00

0.25
Weighted Mean TE Event Rate

0.20

0.15

0.10

0.05

0.00

RCT Comp Noncomp RCT Comp Noncomp RCT Comp Noncomp RCT Comp Noncomp RCT Comp Noncomp RCT Comp Noncomp
Obs Obs Obs Obs Obs Obs Obs Obs Obs Obs Obs Obs

Harms analyses include patients who received recombinant factor VIIa from registries and cohorts (noncomparative observational studies) with at least
15 patients (ICH [70 –72], cardiac surgery [73– 83], body trauma [84 – 87], brain trauma [88, 89], and liver transplantation [90, 91]), as well as patients
from the treatment groups of all RCTs (30 – 43) and comparative observational studies (44 – 69), regardless of quality. For liver transplantation, the
reported RCT rate of TE events is an underestimate, because 1 RCT (40) did not report venous events by group (treatment vs. placebo), so the events
could not be tallied. For studies with overlapping data sets (e.g., the same registry patients being evaluated in a noncomparative study and a comparative
observational study), the most complete data set for the outcome of interest was used. Comp Obs ⫽ comparative observational study; ICH ⫽ intracranial
hemorrhage; Noncomp Obs ⫽ noncomparative observational study; RCT ⫽ randomized, controlled trial; TE ⫽ thromboembolic.

strength is too limited to compare the harms and benefits
of rFVIIa.
Prostatectomy
There was 1 RCT on prophylactic use of rFVIIa in 24
patients having retropubic prostatectomy for prostate can-
cer or benign prostatic hypertrophy (92) (providing an in-
sufficient strength of evidence for all outcomes). Mortality
and thromboembolic events could not be evaluated be-
cause of limited events (0 deaths, 1 thromboembolic event)
(Appendix Table 11). Red blood cell transfusion require-
ments and operating room time were significantly de-
creased with rFVIIa use (Appendix Table 12).
Evaluation of Harms: Data From RCTs Versus
Comparative and Noncomparative Observational Studies
Unadjusted mortality rates among patients who re-
ceived rFVIIa ranged widely from 0 to 0.87 (median,
0.17), and thromboembolism rates ranged from 0 to 0.39
(median, 0.09) (6). In general, mortality rates were lowest
in RCTs and highest in observational studies, but the re-
lationship between study type and thromboembolism rate
was less clear (Figure 4). Randomized, controlled trials of
ICH—which had the longest follow-up and most com-
pletely described ascertainment of harms— had mortality
536 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8
and thrombembolic event rates only slightly different from
those in observational studies. The discrepancy between
harms reported in RCTs versus observational studies was
greater for other indications. There was no apparent corre-
lation between rFVIIa dose and harms outcomes (6),
which is similar to our effectiveness review finding for ICH
of no dose effect. There also was no apparent relationship
between age and harms outcomes (6), although it is nota-
ble that ages tended to cluster for a given indication, which
may have made any relationship difficult to ascertain. Fi-
nally, there was no observable pattern between the likeli-
hood of mortality and thromboembolism (6), which is
similar to the effectiveness review findings for certain indi-
cations (ICH and cardiac surgery) of no effect of rFVIIa on
mortality, despite an associated increase in thromboem-
bolic events.
DISCUSSION
Available evidence indicates that in-hospital, off-label
use of rFVIIa does not reduce mortality for any indication
we evaluated. In contrast, use of rFVIIa increases the rate
of thromboembolic events in ICH and cardiac surgery. Of
the indications studied, the benefit–risk ratio may be most
favorable for body trauma because rFVIIa use in this pop-
www.annals.org
 Effectiveness of Off-Label Use of Recombinant Factor VIIa
ulation is associated with a reduced risk for ARDS, al-
though this finding must be placed in the context of no
accompanying reduction in mortality. In addition, the
strength of evidence is low or moderate for all outcomes,
which precludes definitive conclusions. The harms analyses
raise additional concerns by noting that mortality rates
among patients receiving rFVIIa are generally greater in
observational studies—in which patient groups more
closely reflect unselected patient populations— compared
with RCTs. Our analysis of data from nonfederal U.S.
hospitals found that 97% of in-hospital use of rFVIIa was
off-label, with most for ICH, cardiac surgery, and trauma
indications (93). This increasingly frequent off-label use of
rFVIIa is occurring despite its expense (at least $10 000 for
a single 90-mcg/kg dose in a patient weighing 70 kg) (94),
lack of mortality benefit, and growing evidence of associ-
ated harms.
Our systematic review in some cases extends the find-
ings of previous reviews by incorporating data from com-
parative observational studies in the effectiveness review,
assessing the effect of dosing level when possible and as-
sessing harms across different study types known to have
varying degrees of generalizability to real-world popula-
tions. We are able to provide indication-specific findings
for disparate clinical applications, as well as an overview of
use that identifies outcome patterns across indications,
such as an increase in thromboembolism for some. Previ-
ous reviews that pooled data for several indications identi-
fied no mortality benefit with off-label rFVIIa use (2, 95–
100), and many reviews noted a nonsignificant increase in
thromboembolism (2, 96) (98 –100) or, more recently, a
significant increase (5), but none could provide indication-
specific guidance. Regarding specific indications, a Co-
chrane review of ICH and hemostatic drug therapies, over-
whelmingly rFVIIa, noted no mortality reduction but a
nonsignificant increase in thromboembolism (3), whereas a
more recent review by Yuan and colleagues (101) had the
same findings as ours regarding a significantly increased
risk for arterial thromboembolism without any attendant
benefits for mortality or functional outcome. A meta-
analysis of cardiac surgery observed no effect on mortality
and a nonsignificant increase in perioperative stroke (4).
For body trauma, 3 systematic reviews concluded that
rFVIIa did not decrease mortality (102–104). For brain
trauma, a recent Cochrane review by Perel and coworkers
(105) identified few studies and no mortality benefit.
Other reviews were more supportive of rFVIIa, describing
it as a “promising” therapy (106) or one that might rea-
sonably be used as rescue therapy (107, 108) but did not
do meta-analyses. Specific to harms analyses, O’Connell
and colleagues (1) evaluated data from the FDA’s Adverse
Event Reporting System to document serious thromboem-
bolic events after off-label rFVIIa use.
Our analysis has several limitations. First, the hetero-
geneity of included studies prevented us from pooling data
across clinical indications. Second, we found no studies
www.annals.org
Review
comparing rFVIIa with predecessor products, only with
placebo or usual care. Comparisons with usual care may be
particularly susceptible to bias from site-to-site variations
or advances in clinical care over time. Third, we excluded
articles not published in English, albeit only 2 were com-
parative studies with very few patients. Fourth, we used
quality criteria to determine which observational studies
were assessed in detail in the effectiveness review, such that
others may have come to different determinations. None-
theless, these criteria were applied systematically, and qual-
itative sensitivity testing found that the poor-quality stud-
ies had similar findings to their better-quality counterparts.
Fifth, in 2 of the body trauma RCTs (36), patients who
had been enrolled under protocols with emergency excep-
tions to informed consent withdrew consent. Because no
information was provided on their treatment groups, we
were unable to assess for differential withdrawal. Sixth, the
manufacturer of rFVIIa, Novo Nordisk, played a substan-
tial role in sponsoring, designing, directing, analyzing, and
publishing much of the RCT evidence. Although this cir-
cumstance might be expected and does not equate with
biased research, it does require special care in evaluating
the possibility of bias. Next, we cannot exclude the possi-
bility of selective reporting or publication bias, because the
small number of available studies prevents their formal as-
sessments (for example, by using funnel plots). Finally, we
found that available studies used usual care as the compar-
ator and relied heavily on indirect outcomes, both of which
may yield more favorable findings than other design op-
tions, such as those that use an active comparator or direct
outcomes.
In conclusion, off-label use of rFVIIa for ICH and
cardiac surgery does not reduce mortality but does increase
the risk for thromboembolism. For body trauma, there was
no increased risk for thromboembolism and a reduced risk
for ARDS but no difference in mortality. For the remain-
ing indications, the available evidence was too limited to
do meta-analyses. Thus, limited evidence on off-label
rFVIIa use for 5 indications detects no mortality benefit
but does detect an increase in thromboembolism for some
indications.
From Stanford School of Medicine and Stanford University, Stanford,
California; Palo Alto Medical Foundation Research Institute and Veter-
ans Affairs Palo Alto Health Care System, Palo Alto, California; Univer-
sity of Pennsylvania School of Medicine, Philadelphia, Pennsylvania; and
Castlight Health, San Francisco, California.
Disclaimer: The authors of this report are responsible for its content.
Statements in the report should not be construed as endorsement by the
Agency for Healthcare Research and Quality, the U.S. Department of
Health and Human Services, or the U.S. Department of Veterans Affairs.
Acknowledgment: The authors are grateful to David Johnston, Chris-
topher D. Stave, and James L. Zehnder of Stanford University for their
contributions to this effectiveness review.
19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 537
Review Effectiveness of Off-Label Use of Recombinant Factor VIIa
Grant Support: By the Agency for Healthcare Research and Quality,
U.S. Department of Health and Human Services, under contract 290-
02-0017. Dr. Owens was supported by the U.S. Department of Veterans
Affairs. This research also was supported by training grants from the Na-
tional Heart, Lung, and Blood Institute (RSS, K24HL086703) and the Palo
Alto Medical Foundation Research Institute.
Potential Conflicts of Interest: Drs. Yank, Logan, and Bravata; Ms.
Sundaram; and Ms. McDonald: Grant (money to institution): Agency for
Healthcare Research and Quality. Dr. Staudenmayer: Employment: Stan-
ford Hospital. Dr. Owens: Support for travel to meetings for the study or
other purposes: Agency for Healthcare Research and Quality; Grant
(money to institution): Agency for Healthcare Research and Quality; Con-
sultancy: sanofi-aventis, Generation Health. Dr. Stafford: Expert testi-
mony: Mylan Pharmaceuticals. Disclosures can also be viewed at www
.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum⫽M10
-2333.
Requests for Single Reprints: Veronica Yank, MD, Stanford Preven-
tion Research Center, Stanford University, Medical School Office Build-
ing X312, 251 Campus Drive, Stanford, CA 94304-5411; e-mail, vyank
@stanford.edu.
Current author addresses and author contributions are available at www
.annals.org.
References
1. O’Connell KA, Wood JJ, Wise RP, Lozier JN, Braun MM. Thromboem-
bolic adverse events after use of recombinant human coagulation factor VIIa.
JAMA. 2006;295:293-8. [PMID: 16418464]
2. Stanworth SJ, Birchall J, Doree CJ, Hyde C. Recombinant factor VIIa for the
prevention and treatment of bleeding in patients without haemophilia. Cochrane
Database Syst Rev. 2007:CD005011. [PMID: 17443565]
3. You H, Al-Shahi R. Haemostatic drug therapies for acute primary intracere-
bral haemorrhage. Cochrane Database Syst Rev. 2006;3:CD005951. [PMID:
16856109]
4. Zangrillo A, Mizzi A, Biondi-Zoccai G, Bignami E, Calabrò MG, Pappa-
lardo F, et al. Recombinant activated factor VII in cardiac surgery: a meta-
analysis. J Cardiothorac Vasc Anesth. 2009;23:34-40. [PMID: 19081268]
5. Levi M, Levy JH, Andersen HF, Truloff D. Safety of recombinant activated
factor VII in randomized clinical trials. N Engl J Med. 2010;363:1791-800.
[PMID: 21047223]
6. Yank V, Tuohy CV, Logan AC, Bravata DM, Staudenmayer K, Eisenhut R,
et al. Comparative effectiveness of recombinant factor VIIa for off-label indica-
tions versus usual care. Prepared by Stanford-UCSF Evidence-based Practice
Center under contract no. 290-02-0017. Rockville, MD: Agency for Healthcare
Research and Quality; 2011. Accessed at ww.effectivehealthcare.ahrq.gov\\reports
\\final.cfm on 1 March 2011.
7. Ma B, Wang Z, Zhang B. Effect of recombinant activated factor VII a on
early recovery of patients undergoing cardiac valve replacement under cardiopul-
monary bypass: a randomized double-blind placeb-controlled trial. Dier Junyi
Daxue Xuebao. 2006;27:1110-3.
8. Novak V, Mitov L, Rancić Z, Petrović B, Novak M. [The use of recombinant
activated factor VII in traumatic intracranial haemorrhage]. Srp Arh Celok Lek.
2008;136 Suppl 3:193-8. [PMID: 19562867]
9. Dementyeva II, Gladysheva VG, Charnaya MA, Morozov YA. Influence of
recombinant activated factor VIIa thrombocytic hemostasis. Gematologiya i
Transfuziologiya. 2005;50:12-5.
10. Dementyeva I, Sandrikov V, Charnaya M, Morozov Y, Trekova N, Ere-
menko A. The “NovoSeven” (rFVIIa) hemostatic in massive hemorrhages in
patients operated on the heart and aorta. Anesteziologiya i Reanimatologiya.
2004;5:11-3.
11. Garcia E, Nikolic DS, Piguet V. HIV-1 replication in dendritic cells occurs
through a tetraspanin-containing compartment enriched in AP-3. Traffic. 2008;
9:200-14. [PMID: 18034776]
12. Gong ZY, Gao CQ, Xiao CS, Li BJ, Ma XH, Zhang CM. [The use of
538 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8
recombinant activated factor VII for blood loss after cardiovascular surgery].
Zhonghua Wai Ke Za Zhi. 2008;46:1497-501. [PMID: 19094631]
13. Millán C, Quintana B, Rodrı́guez A, Iglesias M, Barranco M, Navia J.
[Efficacy of recombinant activated factor VII for massive bleeding after cardiac
surgery: experience with 32 patients]. Rev Esp Anestesiol Reanim. 2009;56:485-
92. [PMID: 19994617]
14. Grintescu I, Tulbure D, Mirea L. [Activated recombinant factor VII (No-
voseven) in multiple trauma patients: an outcome analysis]. Chirurgia (Bucur).
2006;101:615-24. [PMID: 17283837]
15. Vandenbroucke JP, Psaty BM. Benefits and risks of drug treatments: how to
combine the best evidence on benefits with the best data about adverse effects.
JAMA. 2008;300:2417-9. [PMID: 19033592]
16. Agency for Healthcare Research and Quality. Methods Reference Guide for
Effectiveness and Comparative Effectiveness Reviews, Version 1.0 [Draft posted
October 2007]. Rockville, MD: Accessed at Agency for Healthcare Research
and Quality; 2007. Accessed at http://effectivehealthcare.ahrq.gov/repFiles
/2007_10DraftMethodsGuide.pdf on 11 March 2011.
17. Owens DK, Lohr KN, Atkins D, Treadwell JR, Reston JT, Bass EB, et al.
AHRQ series paper 5: grading the strength of a body of evidence when compar-
ing medical interventions—agency for healthcare research and quality and the
effective health-care program. J Clin Epidemiol. 2010;63:513-23. [PMID:
19595577]
18. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ,
et al. Assessing the quality of reports of randomized clinical trials: is blinding
necessary? Control Clin Trials. 1996;17:1-12. [PMID: 8721797]
19. Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias.
Dimensions of methodological quality associated with estimates of treatment ef-
fects in controlled trials. JAMA. 1995;273:408-12. [PMID: 7823387]
20. Moher D, Pham B, Jones A, Cook DJ, Jadad AR, Moher M, et al. Does
quality of reports of randomised trials affect estimates of intervention efficacy
reported in meta-analyses? Lancet. 1998;352:609-13. [PMID: 9746022]
21. Kjaergard LL, Villumsen J, Gluud C. Reported methodologic quality and
discrepancies between large and small randomized trials in meta-analyses. Ann
Intern Med. 2001;135:982-9. [PMID: 11730399]
22. Moher D, Schulz KF, Altman DG; CONSORT GROUP (Consolidated
Standards of Reporting Trials). The CONSORT statement: revised recommen-
dations for improving the quality of reports of parallel-group randomized trials.
Ann Intern Med. 2001;134:657-62. [PMID: 11304106]
23. Sanderson S, Tatt ID, Higgins JP. Tools for assessing quality and suscepti-
bility to bias in observational studies in epidemiology: a systematic review and
annotated bibliography. Int J Epidemiol. 2007;36:666-76. [PMID: 17470488]
24. Deeks JJ, Dinnes J, D’Amico R, Sowden AJ, Sakarovitch C, Song F, et al;
International Stroke Trial Collaborative Group. Evaluating non-randomised
intervention studies. Health Technol Assess. 2003;7:iii-x, 1-173. [PMID:
14499048]
25. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenb-
roucke JP; STROBE Initiative. The Strengthening the Reporting of Observa-
tional Studies in Epidemiology (STROBE) statement: guidelines for reporting
observational studies. PLoS Med. 2007;4:e296. [PMID: 17941714]
26. Hedges L, Olkin I. General linear models. In: Statistical Methods for Meta-
analysis. New York: Academic Pr; 1985:166-88.
27. Kranzler HR, Van Kirk J. Efficacy of naltrexone and acamprosate for alco-
holism treatment: a meta-analysis. Alcohol Clin Exp Res. 2001;25:1335-41.
[PMID: 11584154]
28. Gleser L, Olkin I. Stochastically dependent effect sizes. In: Cooper J, Hedges
L, Valentine J, eds. The Handbook of Research Synthesis. New York: Russell
Sage Foundation; 2009:357-76.
29. Rücker G, Schwarzer G, Carpenter J, Olkin I. Why add anything to noth-
ing? The arcsine difference as a measure of treatment effect in meta-analysis with
zero cells. Stat Med. 2009;28:721-38. [PMID: 19072749]
30. Mayer SA, Brun NC, Begtrup K, Broderick J, Davis S, Diringer MN, et al;
Recombinant Activated Factor VII Intracerebral Hemorrhage Trial Investiga-
tors. Recombinant activated factor VII for acute intracerebral hemorrhage. N
Engl J Med. 2005;352:777-85. [PMID: 15728810]
31. Mayer SA, Brun NC, Broderick J, Davis S, Diringer MN, Skolnick BE,
et al; Europe/AustralAsia NovoSeven ICH Trial Investigators. Safety and feasi-
bility of recombinant factor VIIa for acute intracerebral hemorrhage. Stroke.
2005;36:74-9. [PMID: 15569871]
32. Mayer SA, Brun NC, Broderick J, Davis SM, Diringer MN, Skolnick BE,
et al; United States NovoSeven ICH Trial Investigators. Recombinant activated
www.annals.org
 Effectiveness of Off-Label Use of Recombinant Factor VIIa
factor VII for acute intracerebral hemorrhage: US phase IIA trial. Neurocrit Care.
2006;4:206-14. [PMID: 16757825]
33. Mayer SA, Brun NC, Begtrup K, Broderick J, Davis S, Diringer MN, et al;
FAST Trial Investigators. Efficacy and safety of recombinant activated factor VII
for acute intracerebral hemorrhage. N Engl J Med. 2008;358:2127-37. [PMID:
18480205]
34. Diprose P, Herbertson MJ, O’Shaughnessy D, Gill RS. Activated recombi-
nant factor VII after cardiopulmonary bypass reduces allogeneic transfusion in
complex non-coronary cardiac surgery: randomized double-blind placebo-
controlled pilot study. Br J Anaesth. 2005;95:596-602. [PMID: 16183679]
35. Gill R, Herbertson M, Vuylsteke A, Olsen PS, von Heymann C, Mythen
M, et al. Safety and efficacy of recombinant activated factor VII: a randomized
placebo-controlled trial in the setting of bleeding after cardiac surgery. Circula-
tion. 2009;120:21-7. [PMID: 19546387]
36. Boffard KD, Riou B, Warren B, Choong PI, Rizoli S, Rossaint R, et al;
NovoSeven Trauma Study Group. Recombinant factor VIIa as adjunctive ther-
apy for bleeding control in severely injured trauma patients: two parallel random-
ized, placebo-controlled, double-blind clinical trials. J Trauma. 2005;59:8-15.
[PMID: 16096533]
37. Hauser CJ, Boffard K, Dutton R, Bernard GR, Croce MA, Holcomb JB,
et al; CONTROL Study Group. Results of the CONTROL trial: efficacy and
safety of recombinant activated factor VII in the management of refractory trau-
matic hemorrhage. J Trauma. 2010;69:489-500. [PMID: 20838118]
38. Narayan RK, Maas AI, Marshall LF, Servadei F, Skolnick BE, Tillinger
MN; rFVIIa Traumatic ICH Study Group. Recombinant factor VIIA in trau-
matic intracerebral hemorrhage: results of a dose-escalation clinical trial. Neuro-
surgery. 2008;62:776-86. [PMID: 18496183]
39. Lodge JP, Jonas S, Jones RM, Olausson M, Mir-Pallardo J, Soefelt S, et al;
rFVIIa OLT Study Group. Efficacy and safety of repeated perioperative doses of
recombinant factor VIIa in liver transplantation. Liver Transpl. 2005;11:973-9.
[PMID: 16035095]
40. Planinsic RM, van der Meer J, Testa G, Grande L, Candela A, Porte RJ,
et al. Safety and efficacy of a single bolus administration of recombinant factor
VIIa in liver transplantation due to chronic liver disease. Liver Transpl. 2005;11:
895-900. [PMID: 16035081]
41. Pugliese F, Ruberto F, Summonti D, Perrella S, Cappannoli A, Tosi A,
et al. Activated recombinant factor VII in orthotopic liver transplantation. Trans-
plant Proc. 2007;39:1883-5. [PMID: 17692642]
42. Liu J, Chen T, Wang J, AL E. Hemorrhagic features and the application of
recombinant activated factor VII in liver transplantation. Journal of Clinical Re-
habilitative Tissue Engineering Research. 2009;13:3580-4.
43. Friederich PW, Henny CP, Messelink EJ, Geerdink MG, Keller T, Kurth
KH, et al. Effect of recombinant activated factor VII on perioperative blood loss
in patients undergoing retropubic prostatectomy: a double-blind placebo-
controlled randomised trial. Lancet. 2003;361:201-5. [PMID: 12547542]
44. Ilyas C, Beyer GM, Dutton RP, Scalea TM, Hess JR. Recombinant factor
VIIa for warfarin-associated intracranial bleeding. J Clin Anesth. 2008;20:276-9.
[PMID: 18617125]
45. Karkouti K, Beattie WS, Wijeysundera DN, Yau TM, McCluskey SA,
Ghannam M, et al. Recombinant factor VIIa for intractable blood loss after
cardiac surgery: a propensity score-matched case-control analysis. Transfusion.
2005;45:26-34. [PMID: 15647015]
46. Gelsomino S, Lorusso R, Romagnoli S, Bevilacqua S, De Cicco G, Billè G,
et al. Treatment of refractory bleeding after cardiac operations with low-dose
recombinant activated factor VII (NovoSeven): a propensity score analysis. Eur J
Cardiothorac Surg. 2008;33:64-71. [PMID: 17996457]
47. von Heymann C, Redlich U, Jain U, Kastrup M, Schroeder T, Sander M,
et al. Recombinant activated factor VII for refractory bleeding after cardiac
surgery—a retrospective analysis of safety and efficacy. Crit Care Med. 2005;33:
2241-6. [PMID: 16215377]
48. Tritapepe L, De Santis V, Vitale D, Nencini C, Pellegrini F, Landoni G,
et al. Recombinant activated factor VII for refractory bleeding after acute aortic
dissection surgery: a propensity score analysis. Crit Care Med. 2007;35:1685-90.
[PMID: 17522585]
49. Rizoli SB, Nascimento B Jr, Osman F, Netto FS, Kiss A, Callum J, et al.
Recombinant activated coagulation factor VII and bleeding trauma patients. J
Trauma. 2006;61:1419-25. [PMID: 17159685]
50. Fox CJ, Mehta SG, Cox ED, Kragh JF Jr, Salinas J, Holcomb JB. Effect of
recombinant factor VIIa as an adjunctive therapy in damage control for wartime
vascular injuries: a case control study. J Trauma. 2009;66:S112-9. [PMID:
www.annals.org
Review
19359954]
51. Wade CE, Eastridge BJ, Jones JA, West SA, Spinella PC, Perkins JG, et al.
Use of recombinant factor VIIa in US military casualties for a five-year period. J
Trauma. 2010;69:353-9. [PMID: 20699744]
52. Stein DM, Dutton RP, Kramer ME, Handley C, Scalea TM. Recombinant
factor VIIa: decreasing time to intervention in coagulopathic patients with severe
traumatic brain injury. J Trauma. 2008;64:620-7. [PMID: 18332801]
53. Hendriks HG, Meijer K, de Wolf JT, Klompmaker IJ, Porte RJ, de Kam
PJ, et al. Reduced transfusion requirements by recombinant factor VIIa in ortho-
topic liver transplantation: a pilot study. Transplantation. 2001;71:402-5.
[PMID: 11233901]
54. Pickard JD, Kirkpatrick PJ, Melsen T, Andreasen RB, Gelling L, Fryer T,
et al. Potential role of NovoSeven in the prevention of rebleeding following
aneurysmal subarachnoid haemorrhage. Blood Coagul Fibrinolysis. 2000;11
Suppl 1:S117-20. [PMID: 10850576]
55. Brody DL, Aiyagari V, Shackleford AM, Diringer MN. Use of recombinant
factor VIIa in patients with warfarin-associated intracranial hemorrhage. Neuro-
crit Care. 2005;2:263-7. [PMID: 16159073]
56. Hallevi H, Gonzales NR, Barreto AD, Martin-Schild S, Albright KC,
Noser EA, et al. The effect of activated factor VII for intracerebral hemorrhage
beyond 3 hours versus within 3 hours [Letter]. Stroke. 2008;39:473-5. [PMID:
18174476]
57. Bowman LJ, Uber WE, Stroud MR, Christiansen LR, Lazarchick J, Crum-
bley AJ 3rd, et al. Use of recombinant activated factor VII concentrate to control
postoperative hemorrhage in complex cardiovascular surgery. Ann Thorac Surg.
2008;85:1669-76. [PMID: 18442563]
58. Trowbridge C, Stammers A, Klayman M, Brindisi N, Woods E. Charac-
teristics of uncontrolled hemorrhage in cardiac surgery. J Extra Corpor Technol.
2008;40:89-93. [PMID: 18705543]
59. Mitra B, Phillips L, Cameron PA, Billah B, Reid C. The safety of recom-
binant factor VIIa in cardiac surgery. Anaesth Intensive Care. 2010;38:671-7.
[PMID: 20715730]
60. Dutton RP, McCunn M, Hyder M, D’Angelo M, O’Connor J, Hess JR,
et al. Factor VIIa for correction of traumatic coagulopathy. J Trauma. 2004;57:
709-18. [PMID: 15514523]
61. Harrison TD, Laskosky J, Jazaeri O, Pasquale MD, Cipolle M. “Low-dose”
recombinant activated factor VII results in less blood and blood product use in
traumatic hemorrhage. J Trauma. 2005;59:150-4. [PMID: 16096555]
62. Woodruff SI, Dougherty AL, Dye JL, Mohrle CR, Galarneau MR. Use of
recombinant factor VIIA for control of combat-related haemorrhage. Emerg Med
J. 2010;27:121-4. [PMID: 20156864]
63. Tawil I, Stein DM, Mirvis SE, Scalea TM. Posttraumatic cerebral infarction:
incidence, outcome, and risk factors. J Trauma. 2008;64:849-53. [PMID:
18404047]
64. Howes JL, Smith RS, Helmer SD, Taylor SM. Complications of recombi-
nant activated human coagulation factor VII. Am J Surg. 2009;198:895-9.
[PMID: 19969148]
65. Brown CV, Foulkrod KH, Lopez D, Stokes J, Villareal J, Foarde K, et al.
Recombinant factor VIIa for the correction of coagulopathy before emergent
craniotomy in blunt trauma patients. J Trauma. 2010;68:348-52. [PMID:
20154547]
66. Nishijima DK, Dager WE, Schrot RJ, Holmes JF. The efficacy of factor
VIIa in emergency department patients with warfarin use and traumatic intracra-
nial hemorrhage. Acad Emerg Med. 2010;17:244-51. [PMID: 20370756]
67. Kaliciński P, Markiewicz M, Kamiński A, Laniewski P, Ismail H,
Drewniak T, et al. Single pretransplant bolus of recombinant activated factor VII
ameliorates influence of risk factors for blood loss during orthotopic liver trans-
plantation. Pediatr Transplant. 2005;9:299-304. [PMID: 15910384]
68. Niemann CU, Behrends M, Quan D, Eilers H, Gropper MA, Roberts JP,
et al. Recombinant factor VIIa reduces transfusion requirements in liver trans-
plant patients with high MELD scores. Transfus Med. 2006;16:93-100. [PMID:
16623915]
69. De Gasperi A, Baudo F, De Carlis L. Recombinant FVII in orthotopic liver
transplantation (OLT): a preliminary single centre experience [Letter]. Intensive
Care Med. 2005;31:315-6. [PMID: 15592812]
70. Sutherland CS, Hill MD, Kaufmann AM, Silvaggio JA, Demchuk AM,
Sutherland GR. Recombinant factor VIIa plus surgery for intracerebral hemor-
rhage. Can J Neurol Sci. 2008;35:567-72. [PMID: 19235439]
71. Nussbaum ES, Janjua TM, Defillo A, Sinner P, Zelensky A. Perioperative
use of recombinant factor VII to prevent intraoperative aneurysm rupture in high
19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 539
Review Effectiveness of Off-Label Use of Recombinant Factor VIIa
risk patients: a preliminary safety evaluation. Neurocrit Care. 2009;10:55-60.
[PMID: 18818887]
72. Robinson MT, Rabinstein AA, Meschia JF, Freeman WD. Safety of recom-
binant activated factor VII in patients with warfarin-associated hemorrhages of
the central nervous system. Stroke. 2010;41:1459-63. [PMID: 20522813]
73. Tatoulis J, Theodore S, Meswani M, Wynne R, Hon-Yap C, Powar N. Safe
use of recombinant activated factor VIIa for recalcitrant postoperative haemor-
rhage in cardiac surgery. Interact Cardiovasc Thorac Surg. 2009;9:459-62.
[PMID: 19542087]
74. Filsoufi F, Castillo JG, Rahmanian PB, Scurlock C, Fischer G, Adams DH.
Effective management of refractory postcardiotomy bleeding with the use of re-
combinant activated factor VII. Ann Thorac Surg. 2006;82:1779-83. [PMID:
17062247]
75. Gandhi MJ, Pierce RA, Zhang L, Moon MR, Despotis GJ, Moazami N.
Use of activated recombinant factor VII for severe coagulopathy post ventricular
assist device or orthotopic heart transplant. J Cardiothorac Surg. 2007;2:32.
[PMID: 17617902]
76. Hyllner M, Houltz E, Jeppsson A. Recombinant activated factor VII in the
management of life-threatening bleeding in cardiac surgery. Eur J Cardiothorac
Surg. 2005;28:254-8. [PMID: 15935687]
77. McCall P, Story DA, Karalapillai D, Karapillai D. Audit of factor VIIa for
bleeding resistant to conventional therapy following complex cardiac surgery. Can
J Anaesth. 2006;53:926-33. [PMID: 16960271]
78. Raivio P, Suojaranta-Ylinen R, Kuitunen AH. Recombinant factor VIIa in
the treatment of postoperative hemorrhage after cardiac surgery. Ann Thorac
Surg. 2005;80:66-71. [PMID: 15975342]
79. Aggarwal A, Malkovska V, Catlett JP, Alcorn K. Recombinant activated
factor VII (rFVIIa) as salvage treatment for intractable hemorrhage. Thromb J.
2004;2:9. [PMID: 15530167]
80. Karkouti K, Beattie WS, Arellano R, Aye T, Bussieres JS, Callum JL, et al.
Comprehensive Canadian review of the off-label use of recombinant activated
factor VII in cardiac surgery. Circulation. 2008;118:331-8. [PMID: 18606914]
81. Bruckner BA, DiBardino DJ, Ning Q, Adeboygeun A, Mahmoud K,
Valdes J, et al. High incidence of thromboembolic events in left ventricular assist
device patients treated with recombinant activated factor VII. J Heart Lung
Transplant. 2009;28:785-90. [PMID: 19632574]
82. Masud F, Bostan F, Chi E, Pass SE, Samir H, Stuebing K, et al. Recom-
binant factor VIIa treatment of severe bleeding in cardiac surgery patients: a
retrospective analysis of dosing, efficacy, and safety outcomes. J Cardiothorac
Vasc Anesth. 2009;23:28-33. [PMID: 18948033]
83. Hsia CC, Zurawska JH, Tong MZ, Eckert K, McAlister VC, Chin-Yee IH.
Recombinant activated factor VII in the treatment of non-haemophilia patients:
physician under-reporting of thromboembolic adverse events. Transfus Med.
2009;19:43-9. [PMID: 19302454]
84. Knudson MM, Cohen MJ, Reidy R, Jaeger S, Bacchetti P, Jin C, et al.
Trauma, transfusions, and use of recombinant factor VIIa: a multicenter case
registry report of 380 patients from the Western Trauma Association. J Am Coll
Surg. 2011;212:87-95. [PMID: 21115374]
85. Cameron P, Phillips L, Balogh Z, Joseph A, Pearce A, Parr M, et al. The
use of recombinant activated factor VII in trauma patients: experience from the
Australian and New Zealand haemostasis registry. Injury. 2007;38:1030-8.
[PMID: 17706654]
86. Felfernig M; European rFVIIa Trauma Study Group. Clinical experience
with recombinant activated factor VII in a series of 45 trauma patients. J R Army
Med Corps. 2007;153:32-9. [PMID: 17575875]
87. Martinowitz U, Michaelson M; Israeli Multidisciplinary rFVIIa Task
Force. Guidelines for the use of recombinant activated factor VII (rFVIIa) in
uncontrolled bleeding: a report by the Israeli Multidisciplinary rFVIIa Task
Force. J Thromb Haemost. 2005;3:640-8. [PMID: 15842347]
88. McQuay N Jr, Cipolla J, Franges EZ, Thompson GE. The use of recom-
binant activated factor VIIa in coagulopathic traumatic brain injuries requiring
emergent craniotomy: is it beneficial? J Neurosurg. 2009;111:666-71. [PMID:
19425887]
89. Bartal C, Freedman J, Bowman K, Cusimano M. Coagulopathic patients
540 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8
with traumatic intracranial bleeding: defining the role of recombinant factor VIIa.
J Trauma. 2007;63:725-32. [PMID: 18089997]
90. Flower O, Phillips LE, Cameron P, Gunn K, Dunkley S, Watts A, et al.
Recombinant activated factor VII in liver patients: a retrospective cohort study
from Australia and New Zealand. Blood Coagul Fibrinolysis. 2010;21:207-15.
[PMID: 20182351]
91. Jarosz K, Czupryńska M, Andrzejewska J, Wasilewicz MP, Post M, Lu-
bikowski J, et al. Administration of a recombinant factor Vlla in patients under-
going liver transplantation for fulminant hepatic failure. Transplant Proc. 2009;
41:3088-90. [PMID: 19857684]
92. Friedrich JO. Recombinant activated factor VII for acute intracerebral hem-
orrhage [Letter]. N Engl J Med. 2005;352:2133-4. [PMID: 15901870]
93. Logan AC, Yank V, Stafford RS. Off-label use of recombinant factor vlla in
U.S. hospitals: analysis of hospital discharge records. Ann Intern Med. 2011;154:
516-22.
94. Physicians’ Desk Reference. 63rd ed. Montvale, NJ: Medical Economics;
2009.
95. Squizzato A, Ageno W. Recombinant activated factor VII as a general hae-
mostatic agent: evidence-based efficacy and safety. Curr Drug Saf. 2007;2:155-
61. [PMID: 18690962]
96. Hsia CC, Chin-Yee IH, McAlister VC. Use of recombinant activated factor
VII in patients without hemophilia: a meta-analysis of randomized control trials.
Ann Surg. 2008;248:61-8. [PMID: 18580208]
97. Pohar S, Tsakonas E, Murphy G, Anderson D, Carney D, Moltzan C, et al.
Recombinant activated factor VII in treatment of hemorrhage unrelated to he-
mophilia: a systematic review and economic evaluation. Technology report num-
ber 118. Ottawa: Canadian Agency for Drugs and Technologies in Health; 2009.
98. Birchall J, Stanworth SJ, Duffy MR, Doree CJ, Hyde C. Evidence for the
use of recombinant factor VIIa in the prevention and treatment of bleeding in
patients without hemophilia. Transfus Med Rev. 2008;22:177-87. [PMID:
18572094]
99. Johansson PI. Off-label use of recombinant factor VIIa for treatment of
haemorrhage: results from randomized clinical trials. Vox Sang. 2008;95:1-7.
[PMID: 18435677]
100. Lin Y, Stanworth S, Birchall J, Doree C, Hyde C. Use of recombinant
factor VIIa for the prevention and treatment of bleeding in patients without
hemophilia: a systematic review and meta-analysis. CMAJ. 2011;183:E9-19.
[PMID: 21078742]
101. Yuan ZH, Jiang JK, Huang WD, Pan J, Zhu JY, Wang JZ. A meta-
analysis of the efficacy and safety of recombinant activated factor VII for patients
with acute intracerebral hemorrhage without hemophilia. J Clin Neurosci. 2010;
17:685-93. [PMID: 20399668]
102. Duchesne JC, Mathew KA, Marr AB, Pinsky MR, Barbeau JM, McSwain
NE. Current evidence based guidelines for factor VIIa use in trauma: the good,
the bad, and the ugly. Am Surg. 2008;74:1159-65. [PMID: 19097529]
103. Patanwala AE. Factor VIIa (recombinant) for acute traumatic hemorrhage.
Am J Health Syst Pharm. 2008;65:1616-23. [PMID: 18714107]
104. Nishijima DK, Zehtabchi S. Evidence-based Emergency Medicine/Criti-
cally Appraised Topic. The efficacy of recombinant activated factor VII in severe
trauma. Ann Emerg Med. 2009;54:737-744.e1. [PMID: 19285753]
105. Perel P, Roberts I, Shakur H, Thinkhamrop B, Phuenpathom N, Yut-
thakasemsunt S. Haemostatic drugs for traumatic brain injury. Cochrane Data-
base Syst Rev. 2010:CD007877. [PMID: 20091656]
106. Warren O, Mandal K, Hadjianastassiou V, Knowlton L, Panesar S, John
K, et al. Recombinant activated factor VII in cardiac surgery: a systematic review.
Ann Thorac Surg. 2007;83:707-14. [PMID: 17258029]
107. Karkouti K, Beattie WS, Crowther MA, Callum JL, Chun R, Fremes SE,
et al. The role of recombinant factor VIIa in on-pump cardiac surgery: proceed-
ings of the Canadian Consensus Conference. Can J Anaesth. 2007;54:573-82.
[PMID: 17602044]
108. Dunning J, Versteegh M, Fabbri A, Pavie A, Kolh P, Lockowandt U, et al;
EACTS Audit and Guidelines Committee. Guideline on antiplatelet and anti-
coagulation management in cardiac surgery. Eur J Cardiothorac Surg. 2008;34:
73-92. [PMID: 18375137]
www.annals.org

Current Author Addresses: Drs. Yank and Stafford: Stanford Preven-
tion Research Center, Stanford University, Medical School Office Build-
ing X312, 251 Campus Drive, Stanford, CA 94304-5411.
Mr. Tuohy: 433 South 45th Street, Philadelphia, PA 19104.
Dr. Logan: Division of Blood and Marrow Transplantation, Stanford
University School of Medicine, 269 West Campus Drive, CCSR 2200,
Stanford, CA 94305.
Dr. Bravata: Castlight Health, 685 Market Street, #300, San Francisco,
CA 94105.
Dr. Staudenmayer: 300 Pasteur Drive, H3980, Stanford, CA 94305.
Ms. Eisenhut: 1027 Amarillo Avenue, Palo Alto, CA 94303.
Ms. Sundaram, Ms. McDonald, and Dr. Owens: Center for Health
Policy, Center for Primary Care and Outcomes Research, Stanford Uni-
versity, 117 Encina Commons, Stanford, CA 94305-6019.
Dr. McMahon: Apartment 8, Summerfield, Irishtown Road, Dublin 4,
Ireland.
Dr. Olkin: 950 Lathrop Place, Stanford, CA 94305.
Author Contributions: Conception and design: V. Yank, C.V. Tuohy,
D.M. Bravata, V. Sundaram, I. Olkin, K.M. McDonald, D.K. Owens,
R.S. Stafford.
Analysis and interpretation of the data: V. Yank, C.V. Tuohy, A.C.
Logan, D.M. Bravata, K. Staudenmayer, R. Eisenhut, V. Sundaram,
D. McMahon, I. Olkin, D.K. Owens, R.S. Stafford.
Drafting of the article: V. Yank, D.M. Bravata, D. McMahon, I. Olkin,
R.S. Stafford.
W-178 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8
Annals of Internal Medicine
Critical revision of the article for important intellectual content:
V. Yank, C.V. Tuohy, A.C. Logan, D.M. Bravata, K. Staudenmayer,
V. Sundaram, I. Olkin, K.M. McDonald, D.K. Owens, R.S. Stafford.
Final approval of the article: V. Yank, C.V. Tuohy, K. Staudenmayer,
V. Sundaram, D. McMahon, I. Olkin, K.M. McDonald, D.K. Owens,
R.S. Stafford.
Provision of study materials or patients: R.S. Stafford.
Statistical expertise: D. McMahon, I. Olkin, D.K. Owens.
Obtaining of funding: I. Olkin, K.M. McDonald, D.K. Owens, R.S.
Stafford.
Administrative, technical, or logistic support: C.V. Tuohy, V. Sundaram,
K.M. McDonald, D.K. Owens, R.S. Stafford.
Collection and assembly of data: V. Yank, C.V. Tuohy, A.C. Logan,
R. Eisenhut, V. Sundaram, R.S. Stafford.
109. Nascimento B Jr, Tien H, Pinto R, et al. Dosage of factor VIIa (rFVIIa)
and mortality in bleeding trauma patients. J Trauma. 2008;64:558.
110. Spinella PC, Perkins JG, McLaughlin DF, Niles SE, Grathwohl KW,
Beekley AC, et al. The effect of recombinant activated factor VII on mortality in
combat-related casualties with severe trauma and massive transfusion. J Trauma.
2008;64:286-93. [PMID: 18301188]
111. Lodge JP, Jonas S, Jones RM, Olausson M, Mir-Pallardo J, Soefelt S, et al;
rFVIIa OLT Study Group. Efficacy and safety of repeated perioperative doses of
recombinant factor VIIa in liver transplantation. Liver Transpl. 2005;11:973-9.
[PMID: 16035095]
112. Friederich PW, Henny CP, Messelink EJ, Geerdink MG, Keller T, Kurth
KH, et al. Effect of recombinant activated factor VII on perioperative blood loss
in patients undergoing retropubic prostatectomy: a double-blind placebo-
controlled randomised trial. Lancet. 2003;361:201-5. [PMID: 12547542]
www.annals.org
 Appendix Figure. Summary of evidence search and selection.

CCTR,
ACP Journal Club,
and DARE
MEDLINE EMBASE citations (n = 304) BIOSIS
citations (n = 3270) citations (n = 4424) citations (n = 3190)

Total Exclusions (n = 4846)

unique citations Not rFVIIa: 3043
(n = 6191) Use in patients who have
hemophilia, FVIIa deficiency, or
other bleeding or clotting
disorders: 895
Use for other indications: 25
Level 1: Review of title and Not used as therapy or for
abstract by 2 independent clinically relevant outcomes: 25
investigators Use in animals or in vitro: 277
No usable data: 1
Ineligible study design: 542
Duplicate publication: 25
Non-English language: 5
Other reason: 8

Citations requiring Exclusions (n = 1283)

full-text review
(n = 1345)
Level 2: Review of full text and
abstraction of included studies by
2 independent investigators
Unique citations
included in analyses
(n = 62)
Not rFVIIa: 20
Use in patients who have
hemophilia, FVIIa deficiency, or
other bleeding or clotting
disorders: 54
Use in pediatric patients: 9
Use for other indications: 17
Not used as therapy or for
clinically relevant outcomes: 7
Use in animals or in vitro: 20
Data combined for several
conditions: 28
No usable data: 21
Small sample size: 196
Ineligible study design: 796
Duplicate publication: 84
Non-English language: 3
Unpublished data: 6
Article not available: 7
Other reason: 15

RCT (n = 14) Comparative observational Noncomparative observational

Systematic review (includes (n = 26) (n = 22)
16 RCTs): 14 Systematic review: 10 Harms analysis: 22
Qualitative sensitivity
testing: 16

CCTR ⫽ Cochrane Central Register of Controlled Trials; DARE ⫽ Database of Abstracts of Reviews of Effects; RCT ⫽ randomized, controlled trial;
rFVIIa ⫽ recombinant factor VIIa.

www.annals.org 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 W-179
 Appendix Table 1. Mortality and Poor Functional Outcome on Modified Rankin Scale Score in Comparative Studies of rFVIIa Use in Intracranial Hemorrhage

Study, Year Study Design and Exact or Mean Sample Size, n* Mean Age (SD) [Range], y Mortality Rate Poor Functional Outcome on Rankin
(Reference) rFVIIa Use (SD) [Range] Scale Score Rate†
rFVIIa Dose,
mcg/kg rFVIIa Group Usual rFVIIa Group Usual Care rFVIIa Group Usual Care P Value‡ rFVIIa Group Usual P Value‡
Care Group Group Care
Group Group

Mayer et al, RCT, treatment 40, 80, 160 108, 92, 103 96 67 (12), 65 (12), 68 (12) 0.176, 0.185, 0.292 0.05, 0.10, 0.546, 0.495, 0.544 0.688 0.02, 0.008, 0.02
2005 (30) 64 (13) 0.194 0.11
All patients – – 0.185 0.292 0.02 0.530 0.688 0.004
Mayer et al, RCT, treatment 10, 20, 40, 80, 6, 6, 6, 6, 6, 11 51 (9), 68 (22), 66 (14) 0, 0, 0.167, 0.182 NR 0, 0.333, 0.667, 0.455 NR
2005 (31) 120, 160 6 68 (16), 58 (11), 0, 0.167, 0.5, 0.5, 0.667
64 (14), 53 (12) 0.167
All patients – – 0.083 0.182 NR 0.444 0.455 NR
Mayer et al, RCT, treatment 5, 20, 40, 80 8, 8, 8, 8 8 72 (10), 60 (15), 67 (13) 0.25, 0.25, 0.125 NR 0.625, 0.375, 0.500 NR
2006 (32) 64 (13), 62 (12) 0.375, 0 0.625, 0.25

W-180 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8

All patients – – 0.219 0.125 NR 0.469 0.500 NR
Mayer et al, RCT, treatment 20, 80 276, 297 268 65 (14), 65 (13) 65 (14) 0.181, 0.209 0.190 0.38, 0.75 0.492, 0.505 0.466 1.0 [0.6–1.6]§,
2008 (33) 1.4 [0.9–2.2]§
All patients – – 0.195 0.190 NR 0.499 0.466 NR
Unweighted sum All RCTs 5–40, 80, 120–160 426, 403, 383 – – 0.181, 0.196, 0.214 0.502, 0.497, 0.548 0.522
115 0.191
All patients – – 0.189 0.214 0.506 0.522
Ilyas et al, Retrospective comparative, [10–100] 24 30 76.5 (11) 76.4 (12.4) 0.21 0.27 0.86 NR NR NR
2008 (44) treatment
Pickard et al, Prospective comparative, 80, 80 ⫹ 3.5 CI/h, 5 5 NR NR NR NR NR NR NR NR
2000 (54)㛳 treatment 80 ⫹ 7.0 CI/h
Brody et al, Retrospective comparative, 4.8 (2.1)¶ 12 15 71 (13) 77 (7) 0.42 0.13 NR ** ** NR
2005 (55)㛳 treatment
Hallevi et al, Retrospective comparative, 40, 80 46 148 60 [38–87] 58 [40–80] 0.13 0.14 NR 0.57 0.54 NR
2008 (56)㛳 treatment

CI ⫽ continuous infusion; NR ⫽ not reported; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* This is the largest sample size reported in each study, which is the correct sample size for mortality. Sample size for poor functional outcome on the modified Rankin scale was smaller in Mayer et al (30) and Mayer et al (33):
108, 91, 103, and 96 in the 40-mcg/kg, 80-mcg/kg, 160-mcg/kg, and usual care groups, respectively, in Mayer et al (30), and 264, 293, and 262 in the 20-mcg/kg, 80-mcg/kg, and usual care groups, respectively, in Mayer et al (33).
† Poor functional outcome defined as a modified Rankin scale score of 4 to 6. Data from Rankin scale scores of 4 to 6 in Mayer et al (33) were derived graphically from their figure 3. The proportion of patients with modified
Rankin scale scores of 5 to 6 was reported as 0.261, 0.287, 0.237 for the 20-mcg/kg, 80-mcg/kg, and usual care groups, respectively. The odds ratio reported for Mayer et al (33) is based on the proportion of patients with modified
Rankin scale scores of 5 to 6.
‡ P values presented are those reported by the individual studies, wherever possible.
§ Odds ratio for poor functional outcomes and 95% CI based on the proportion of patients with modified Rankin scale scores of 5 to 6.
㛳 Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes section
and the Evaluation of Harms section.
¶ Mean dose in mg rather than mcg/kg of body weight.
** Brody et al (55) report a median (range) Glasgow Coma Scale score at discharge by group (15 ⫽ normal, 0 ⫽ dead): rFVIIa, 13.5 (13–15) and usual care, 15 (13–15).

www.annals.org
 www.annals.org
Appendix Table 2. Thromboembolic Events (Arterial and Venous) in Comparative Studies of rFVIIa Use in Intracranial Hemorrhage

Study, Year Study Design and Exact or Mean Sample Size, n* Mean Age (SD) [Range], y Total Thromboembolic Arterial Thromboembolic Venous Thromboembolic
(Reference) rFVIIa Use (SD) [Range] Event Rate† Event Rate† Event Rate†
rFVIIa Dose,
mcg/kg rFVIIa Group Usual rFVIIa Group Usual Care rFVIIa Group Usual Care P Value‡ rFVIIa Group Usual P Value‡ rFVIIa Group Usual P Value‡
Care Group Group Care Care
Group Group Group

Mayer et al, RCT, treatment 40, 80, 160 108, 92, 103 96 67 (12), 65 (12), 68 (12) 0.065, 0.043, 0.021 NR 0.056, 0.022, 0 NR 0.009, 0.022, 0.021 NR
2005 (30) 64 (13) 0.097 0.078 0.019
All patients 0.069 0.021 0.12 0.053 0 0.01 0.017 0.021 NR
Mayer et al, RCT, treatment 10, 20, 40, 80, 6, 6, 6, 6, 6, 11 51 (9), 68 (22), 66 (14) 0, 0.333, 0.091 NR 0, 0.167, 0 NR 0, 0.167, 0, 0.091 NR
2005 (31) 120, 160 6 68 (16), 58 (11), 0.333, 0, 0.333, 0, 0, 0, 0, 0
64 (14), 53 (12) 0, 0 0
All patients 0.111 0.091 NR 0.083 0 NR 0.028 0.091 NR
Mayer et al, RCT, treatment 5, 20, 40, 80 8, 8, 8, 8 8 72 (10), 60 (15), 67 (13) 0, 0.25, 0.125 NR 0, 0, 0, 0 0.125 NR 0, 0.25, 0 NR
2006 (32) 64 (13), 62 (12) 0.125, 0 0.125, 0
All patients 0.094 0.125 NR 0 0.125 NR 0.094 0 NR
Mayer et al, RCT, treatment 20, 80 265, 293 263 65 (14), 65 (13) 65 (14) 0.098, 0.126 0.087 NR 0.057, 0.092 0.046 NR, 0.04 0.042, 0.034 0.042 NR, NR
2008 (33)*
All patients 0.113 0.087 NR 0.075 0.046 NR 0.038 0.042 NR
Unweighted sum All RCTs 5–40, 80, 120–160 415, 399, 378 0.096, 0.103, 0.071 0.058, 0.072, 0.034 0.039, 0.030, 0.037
115 0.087 0.070 0.017
All patients 0.098 0.071 0.066 0.034 0.032 0.037
Ilyas et al, Retrospective comparative, [10–100] 24 30 76.5 (11) 76.4 (12.4) 0.04 0 NR 0.04 0 NR 0 0 NR
2008 (44) treatment
Pickard et al, Prospective comparative, 80, 80 ⫹ 3.5 CI/h, 5 5 NR NR 0.2 0 NR 0.2 0 NR 0 0 NR
2000 (54)§ treatment 7.0 CI/h
Brody et al, Retrospective comparative, 4.8 (2.1)㛳 12 15 71 (13) 77 (7) 0.25 0 NR 0.08 0 NR 0.17 0 NR
2005 (55)§ treatment
Hallevi et al, Retrospective comparative, 40, 80 46 148 60 [38–87] 58 [40–80] 0.15 NR NR 0.09 NR NR 0.07 NR NR
2008 (56)§ treatment

CI ⫽ continuous infusion; NR ⫽ not reported; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* Sample size for thromboembolic events in Mayer et al (33) is smaller than reported for mortality and poor modified Rankin scale score because thromboembolic event rates were calculated on the basis of the “safety population”
(i.e., patients exposed to a study agent) rather than the patients who were randomly assigned.
† Thromboembolic event rates were calculated by dividing the number of thromboembolic events by the sample size, not the number of patients who had thromboembolic events. Therefore, the rates reported here may differ
slightly from the rates reported in each study. The tests of statistical significance presented are those reported by the individual studies and are not based on the thromboembolic event rates reported in this table.
‡ P values presented are those reported by the individual studies, wherever possible.
§ Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes
section and the Evaluation of Harms section.
㛳 Mean dose in mg rather than mcg/kg of body weight.

19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 W-181

 Appendix Table 3. Indirect Outcomes in Comparative Studies of rFVIIa Use in Intracranial Hemorrhage

Study, Year Study Design and Exact or Mean Sample Size, n Mean Age (SD) [Range], y Relative Hematoma Expansion (SD) [95% CI], % Absolute Hematoma Expansion (SD) [95% CI], mL
(Reference) rFVIIa Use (SD) [Range]
rFVIIa Dose, rFVIIa Group Usual rFVIIa Group Usual Care rFVIIa Group Usual Care P Value* rFVIIa Group Usual Care P Value*
mcg/kg Care Group Group Group
Group

Mayer et al, RCT, treatment 40, 80, 160 108, 92, 103 96 67 (12), 65 (12), 68 (12) 16 [4–28]†, 29 [16–44]† 0.07, 0.05, 5.4 [1.7–9.0]†, 8.7 [4.9–12.4]† 0.13, 0.14, 0.008‡,
2005 (30) 64 (13) 14 [2–27]†, 0.02‡, all 4.2 [0.3–8.0]†, all patients:
11 [0–23]† patients: 2.9 [⫺0.8–6.6]† 0.01‡
0.01‡
Mayer et al, RCT, treatment 10, 20, 40, 80, 6, 6, 6, 6, 6, 11 51 (9), 68 (22), 66 (14) Same Same NS NR NR NR

W-182 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8

2005 (31) 120, 160 6 68 (16), 58 (11),
64 (14), 53 (12)
Mayer et al, RCT, treatment 5, 20, 40, 80 8, 8, 8, 8 8 72 (10), 60 (15), 67 (13) 64 (106), 11 (18) NS 13 (19), 10 (21), 2 (4) NS
2006 (32) 64 (13), 62 (12) 79 (172), 1 (2), 1 (2)
37 (51), 4 (9)
Mayer et al, RCT, treatment 20, 80 276, 297 268 65 (14), 65 (13) 65 (14) 18 [13–24], 26 [20–32] 0.09, ⬍0.001 4.9 [2.9–7.0], 7.5 [5.4–9.6] 0.08, 0.009
2008 (33) 11 [6–17] 3.7 [1.7–5.7]
Ilyas et al, Retrospective comparative, [10–100] 24 30 76.5 (11) 76.4 (12.4) NR NR NR NR NR NR
2008 (44) treatment
Pickard et al, Prospective comparative, 80, 80 ⫹ 3.5 CI/h, 5 5 NR NR NR NR NR NR NR NR
2000 (54)§ treatment 80 ⫹ 7.0 CI/h
Brody et al, Retrospective comparative, 4.8 (2.1)㛳 12 15 71 (13) 77 (7) NR NR NR NR NR NR
2005 (55)§ treatment
Hallevi et al, Retrospective comparative, 40, 80 46 148 60 [38–87] 58 [40–80] 11.6 (14.6) NR NR 1.09 (1.37) NR NR
2008 (56)§ treatment

CI ⫽ continuous infusion; NR ⫽ not reported; NS ⫽ not significant; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* P values presented are those reported by the individual studies.
† In Mayer et al (30), 98.3% CI is used.
‡ The comparison was statistically significant according to the prespecified Bonferroni-corrected threshold of P ⫽ 0.0167 in Mayer et al (30).
§ Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes
section and the Evaluation of Harms section.
㛳 Mean dose in mg rather than mcg/kg of body weight.

www.annals.org
 www.annals.org
Appendix Table 4. Mortality and Thromboembolic Events in Comparative Studies of rFVIIa Use in Cardiac Surgery

Study, Year (Reference) Study Design and rFVIIa Exact or Sample Size, n Mean Age (SD) [IQR], y Mortality Rate Thromboembolic Event Rate*
Use Mean (IQR)
rFVIIa Dose, rFVIIa Group Usual Care rFVIIa Group Usual Care rFVIIa Group Usual Care P Value† rFVIIa Group Usual Care P Value†
mcg/kg Group Group Group Group
Diprose et al, 2005 (34) RCT, prophylaxis 90 9‡ 10 63§ [59–66] 69.5§ 0 0.1 1 0.222 0.2 NR
[63.5–
76.5]
Gill et al, 2009 (35) RCT, treatment 40, 80 35, 69 68 68 (12), 62 (16) 0.114, 0.087 0.059 NR 0.086, 0.058 0.015 NR
63 (16)
Karkouti et al, Retrospective comparative, 51.1 51 51 56 59 0.137 0.137 NS 0.157 0.098 NR
2005 (45) treatment
Gelsomino et al, Retrospective comparative, 18 (9–16) 40 40 70.1 (9.2) 73.2 (7.8) 0.05 0.075 ⬎0.9 0.05 0 NR
2008 (46) treatment
von Heymann et al, Retrospective comparative, 89.3 24 24 63.5 63.5 0.333㛳 0.333㛳 1 0 0 NR
2005 (47) treatment
Tritapepe et al, Retrospective comparative, 82.2 23 23 62.4 (9.4) 62.2 (9.1) 0.13 0.13 1 0.087 0.087 NR
2007 (48) treatment
Bowman et al, Retrospective comparative, 100 36 385 58 (15) NR 0.083 0.044–0.047** NR 0.111 0.055 0.15
2008 (57)¶ treatment
Trowbridge et al, Prospective comparative, NR 17 187 70 (9) 67 (11) NR NR NR NR NR NR
2008 (58)¶ treatment
Mitra et al, 2010 (59)¶ Prospective comparative, NR 705 6554 66 68 0.186 0.056 ⬍0.001 0.092 0.104 NR
treatment

IQR ⫽ interquartile range; NR ⫽ not reported; NS ⫽ not significant; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* Thromboembolic event rates were calculated by dividing the number of thromboembolic events by the sample size, not the number of patients who had thromboembolic events. Therefore, the rates reported here may differ
slightly from those reported in each study. The tests of statistical significance presented are those reported by the individual studies and are not based on the thromboembolic event rates reported in this table.
† P values presented are those reported by the individual studies.
‡ Diprose et al (34) excluded 1 patient who violated the study protocol after randomization. This patient was not evaluated for mortality or thromboembolic events.
§ Median.
㛳 The hospital mortality rates at 6 mo were 0.583 in the rFVIIa group and 0.417 in the usual care group.
¶ Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes
section and the Evaluation of Harms section.
** These data were extrapolated as the range of possible mortality for the usual care group, given the mortality rates reported for the rFVIIa group and the group of all patients combined.

19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 W-183

 Appendix Table 5. Indirect Outcomes in Comparative Studies of rFVIIa Use in Adult Cardiac Surgery

Study, Year Study Design and rFVIIa Exact or Sample Size, n Mean Age (SD) [IQR], y Mean RBC Transfusion (SD) [IQR], units Mean ICU Length of Stay (SD) [IQR], d
(Reference) Use Mean (IQR)
rFVIIa Dose, rFVIIa Group Usual rFVIIa Group Usual Care Group rFVIIa Group Usual Care P Value* rFVIIa Group Usual Care P Value*
mcg/kg Care Group Group
Group
Diprose et al, RCT, prophylaxis 90 9 10 63† [59–66] 69.5† [63.5–76.5] 0‡ [0–0.75] 2‡ [1–3.5] 0.11‡ 2.5† [1–3.5] 1† [1–4] 0.43
2005 (34)
Gill et al, RCT, treatment 40, 80 35, 69 68 68 (12), 62 (16) § § § No difference㛳 No difference㛳 NR
2009 (35) 63 (16)
Karkouti et al, Retrospective comparative, 51.1 51 51 56 59 2† [0–3] NR NR 6† [3.5–11.5] 3.5† [1–10] ⬍0.05
2005 (45) treatment
Gelsomino et al, Retrospective comparative, 18 (9–16) 40 40 70.1 (9.2) 73.2 (7.8) 6.5† [4–8.5] 17† [12–19.5] ⬍0.001 6.3† [3.9–7.3] 18.5† [16.2–24] ⬍0.001
2008 (46) treatment

W-184 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8

von Heymann Retrospective comparative, 89.3 24 24 63.5§ 63.5§ 4.0¶ 2.6¶ 0.44 14 [6.5–34.5]** 15 [6.0–29.5]** 0.96
et al, 2005 (47) treatment
Tritapepe et al, Retrospective comparative, 82.2 23 23 62.4 (9.4) 62.2 (9.1) 9.1 (3.0) 5.3 (3.2) ⬍0.001 8.2 (9.0) 10.0 (8.4) 0.08
2007 (48) treatment
Bowman et al, Retrospective comparative, 100 36 385 58 (15) NR 3.1‡‡ NR NR NR NR NR
2008 (57)†† treatment
Trowbridge et al, Prospective comparative, NR 17 187 70 (9) 67 (11) 0.6 (1.2)§§ 5.4 (3.5)§§ ⬍0.001 NR NR NR
2008 (58)†† treatment
Mitra et al, Prospective comparative, NR 705 6554 66 68 NR NR NR 18㛳 㛳 10㛳 㛳 ⬍0.001㛳 㛳
2010 (59)†† treatment

ICU ⫽ intensive care unit; IQR ⫽ interquartile range; NR ⫽ not reported; RBC ⫽ red blood cell; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* P values presented are those reported by the individual studies.
† Median.
‡ Intention-to-treat data were used (rFVIIa group sample size was 10 instead of 9). One patient who violated the study protocol in Diprose et al (34) was excluded from the analysis of mortality, thromboembolic events, and ICU
length of stay but was included in the intention-to-treat analysis of RBC transfusions. If this patient is excluded, median RBC transfusions are 0 units (range, 0 – 0 units) in the rFVIIa group and 2 units (range, 1–3.5 units) in
the usual care group (P ⫽ 0.03).
§ Total blood transfusion volume (not isolated to RBCs): for 40-mcg/kg rFVIIa, 640 mL (IQR, 0 –1920) (P ⫽ 0.047 vs. usual care); for 80-mcg/kg rFVIIa, 500 mL (IQR, 0 –1750) (P ⫽ 0.042 vs. usual care); for usual care,
825 mL (IQR, 326.5–1893).
㛳 There was no difference noted in ICU length of stay between groups in the trial synopsis posted on the manufacturer’s Web site. This outcome was not reported in the published article.
¶ Estimated graphically from figure 2 in von Heymann et al (47).
** Range.
†† Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes
section and the Evaluation of Harms section.
‡‡ Estimated graphically from figure 1 in Bowman et al (57).
§§ Post– cardiopulmonary bypass transfusion without adjustment for timing of rFVIIa administration.
㛳 㛳 Reported as length of stay, which indicates full hospital length of stay, not ICU length of stay.

www.annals.org
 www.annals.org
Appendix Table 6. Mortality, Thromboembolic Events, and ARDS in Comparative Studies of rFVIIa Use in Body Trauma

Study, Year Study Design and rFVIIa Exact or Sample Size, n Mean Age (SD) [IQR], y 30-d Mortality Rate Thromboembolic Event Rate ARDS Rate
(Reference) Use Mean (Range)
rFVIIa Dose, rFVIIa Usual Care rFVIIa Group Usual Care rFVIIa Usual Care P Value* rFVIIa Usual Care P Value* rFVIIa Usual Care P Value*
mcg/kg Group Group Group Group Group Group Group Group Group

Hauser et al, RCT, treatment 400 218 242 39.2 (14.3) 39.9 (14.2) 0.110 0.107 0.93 0.200† 0.140† 0.38† 0.036 0.072 0.08
2010 (37)
(blunt)
Boffard et al, RCT, treatment 400 69 74 33 (13) 35 (13) 0.246 0.297 0.58 0.029 0.041 NR 0.043 0.162 0.03
2005 (36)
(blunt)
Hauser et al, RCT, treatment 400 44 38 33.8 (11.9) 29.4 (10.3) 0.182 0.132 0.40 0.043† 0.100† 0.41† 0 0.075 0.10
2010 (37)
(penetrating)
Boffard et al, RCT, treatment 400 70 64 29 (10) 32 (10) 0.243 0.281 0.69 0.057 0.047 NR 0.057 0.078 0.74
2005 (36)
(penetrating)
Rizoli et al, Retrospective comparative, NR 38 204 36.8 (CI, 41.1 (CI, ‡ ‡ CI, 0.8–7.6‡ NR NR NR NR NR NR
2006 (49) treatment 30.6–43.1) 38.1–44.1)
Nascimento et al, – – 72 256 – – – – – 0.097 0.063 0.31 – – –
2008 (109)§
Wade et al, Retrospective comparative, NR 266 266 24¶ [21–29] 24¶ [21–28] 0.199 0.143 0.08 㛳 㛳 㛳 㛳 㛳 㛳
2010 (51)㛳 treatment
Spinella et al, Retrospective comparative, 120 49 75 NR NR 㛳 㛳 㛳 0.041 0 0.15 0.020 0.040 1.0
2008 (110) treatment
Fox et al, Retrospective comparative, (90–120) 41 12 28 (9) 24 (10) 0.073** 0** 1 0.122 0.083 1 NR NR NR
2009 (50) treatment
Dutton et al, Retrospective comparative, 50, 100 81 32–449‡‡ 41 (21) NR 0.580 0.438, NR, 0.60‡‡ 0 NR NR NR NR NR
2004 (60)†† treatment 0.325‡‡
Harrison et al, Retrospective comparative, 60 29 72 41 (21) 42 (22) 0.414** 0.403§§ NR 0.069㛳 㛳 0.197㛳 㛳 0.2 NR NR NR
2005 (61)†† treatment
Woodruff et al, Retrospective comparative, ¶¶ 22 22 24 NR 0.318 0.318 1 NR NR NR NR NR NR
2010 (62)†† treatment

ARDS ⫽ acute respiratory distress syndrome; IQR ⫽ interquartile range; NR ⫽ not reported; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* P values presented are those reported by the individual studies.
† Thromboembolic event rates were calculated by dividing the number of thromboembolic events by the sample size, not the number of patients who had thromboembolic events. Therefore, the rates reported here may differ
slightly from those reported in each study. The tests of statistical significance presented are those reported by the individual studies and are not based on the thromboembolic event rates reported in this table.
‡ This study generated odds ratios for survival by using a multivariate model, which included independent predictors of in-hospital survival (baseline pH, platelet count, age, head injury, and transfusion during the resuscitation
period) as covariates. Before modeling, the unadjusted in-hospital mortality rates were 0.500 in the rFVIIa group and 0.485 in the usual care group. In the multivariate models, rFVIIa was shown to have no significant effect on
in-hospital survival (odds ratio, 2.5 [95% CI, 0.8 –7.6]) but to improve 24-h survival (odds ratio, 3.4 [CI, 1.2–9.8]).
§ Nascimento et al (109) is an extension of Rizoli et al (49) but with more patients. It also reports data on thromboembolic events not reported in the original publication.
㛳 Wade et al (51) is an extension of Spinella et al (110) using patient data from the U.S. Joint Theater Trauma Registry but only reports on total mortality (not thromboembolic event or ARDS rates) for the propensity-matched
group of patients (n ⫽ 266), so that the earlier data from Spinella et al (110) are used for the thromboembolic event and ARDS rates.
¶ Median.
** 24-h mortality.
†† Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes
section and the Evaluation of Harms section.
‡‡ Dutton et al (60) has several control groups. The range of sample sizes is presented. For purposes of outcome data, rates for the most narrow sample (n ⫽ 32) and most broad sample (n ⫽ 449) are given, respectively.
§§ 28-d mortality.
㛳 㛳 Venous thromboembolic events only.
¶¶ The description from Woodruff et al (62) of mean rFVIIa dose may inaccurate because it reports the dose to be 9.6 mg/kg and not mcg/kg, which would be more than 1 magnitude larger than the doses reported in the body
trauma RCTs or any other report of off-label rFVIIa use.

19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 W-185

 Appendix Table 7. Indirect Outcomes in Comparative Studies of rFVIIa Use in Body Trauma

Study, Year (Reference) Study Design and rFVIIa Exact or Sample Size, n* Mean Age (SD) [IQR], y 24-h RBC Transfusion (SD) [IQR], units
Use Mean (Range)
rFVIIa Dose, rFVIIa Group Usual Care rFVIIa Group Usual Care rFVIIa Group Usual Care P Value†
mcg/kg Group Group Group
Hauser et al, 2010 (37) (blunt) RCT, treatment 400 218 242 39.2 (14.3) 39.9 (14.2) 6.9 (10.4) 8.1 (10.9) 0.04
Boffard et al, 2005 (36) (blunt) RCT, treatment 400 59 33 (13) 35 (13) ‡ ‡ 0.07‡
Hauser et al, 2010 (37) RCT, treatment 400 44 38 33.8 (11.9) 29.4 (10.3) 4.5 (7.3) 6.2 (6.5) 0.11
(penetrating)
Boffard et al, 2005 (36) RCT, treatment 400 57 52 29 (10) 32 (10) ‡ ‡ 0.24‡
(penetrating)
Rizoli et al, 2006 (49) Retrospective comparative, NR 38 204 36.8 [CI, 41.1 [CI, § § §
treatment 30.6–43.1] 38.1–44.1]
Wade et al, 2010 (51)¶ Retrospective comparative, NR 266 266 24㛳 [21–29] 24㛳 [21–28] ¶ ¶ ¶
treatment

W-186 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8

Spinella et al, 2008 (110) Retrospective comparative, 120 49 75 NR NR 16㛳 [13–27] 14㛳 [11–19] 0.02
treatment
Fox et al, 2009 (60) Retrospective comparative, (90–120) 41 12 28 (9) 24 (10) NR NR NR
treatment
Dutton et al, 2004 (60)** Retrospective comparative, 50, 100 81 32–449†† 41 (21) NR †† NR NR
treatment
Harrison et al, 2005 (61)** Retrospective comparative, 60 29 72 41 (21) 42 (22) 18.3‡‡ (7.5) 22‡‡ (9.7) 0.04
treatment
Woodruff et al, 2010 (62)** Retrospective comparative, §§ 22 22 24 NR NR NR NR
treatment

IQR ⫽ interquartile range; NR ⫽ not reported; RBC ⫽ red blood cell; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* The sample size for RBC transfusions is smaller in Boffard et al (36) because patients who died within the 48-h observation period for RBC transfusions were excluded.
† P values presented are those reported by the individual studies.
‡ 48-h RBC transfusions for all patients (including those who died before 48 h) are reported as median and ranges but estimated reduction in transfusion requirement with rFVIIa use as units of RBC (90% CI) are also given:
for blunt trauma, 2.0 (CI, 0.0 – 4.6), and penetrating trauma, 0.2 (CI, ⫺0.9 to 2.4).
§ This study used multivariate modeling to generate odd ratios for survival in the rFVIIa group vs. the control group. The model adjusted for independent predictors of in-hospital survival, which included transfusion during the
resuscitation period. Without adjustment for differences in important baseline covariates, RBC transfusions during the first 24 h were 24.9 units (95% CI, 21.5–28.3) in the rFVIIa group and 14.9 units (CI, 14.0 –15.8) in the
usual care group; P ⬍ 0.001.
㛳 Median.
¶ Wade et al (51) is an extension of Spinella et al (110) using patient data from the U.S. Joint Theater Trauma Registry but only reports on total mortality (not thromboembolic event or acute respiratory distress syndrome rates)
for the propensity-matched group of patients (n ⫽ 266), so that the earlier data from Spinella et al (110) are used for the thromboembolic event and acute respiratory distress syndrome rates.
** Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes
section and the Evaluation of Harms section.
†† Dutton et al (60) has multiple control groups. The range of sample sizes is presented. 24-h transfusion results are only given only for patients who received rFVIIa divided into 2 groups: responders who received 4.5 units (SD,
8.0) and nonresponders who received 22.6 units (SD, 14.7).
‡‡ 72-h RBC transfusions.
§§ The description from Woodruff et al (62) of mean rFVIIa dose may inaccurate because it reports the dose to be 9.6 mg/kg and not mcg/kg, which would be more than 1 magnitude larger than the doses reported in the body
trauma RCTs or any other report of off-label rFVIIa use.

www.annals.org
 www.annals.org
Appendix Table 8. Mortality, Thromboembolic Events, and Absolute Change in Hematoma Volume in Comparative Studies of rFVIIa Use in Brain Trauma

Study, Year Study Design Exact or Sample Size, n Mean Age (SD) [Range], y Mortality Rate Thromboembolic Event Rate* Absolute Change in Hematoma Volume
(Reference) and rFVIIa Use Mean (SD) (SD), mL
[Range]
rFVIIa Dose, rFVIIa Group Usual Care rFVIIa Group Usual Care rFVIIa Group Usual Care P Value† rFVIIa Group Usual Care P Value† rFVIIa Group Usual P Value†
mcg/kg Group Group Group Group Care
Group

Narayan et al, RCT, treatment 40, 80, 120, 12, 11, 14, 36 All treatment: 51.5 51.4 (19.5) 0, 0, 0.071, 0.111 NR 0.167, 0.182, 0.083 NR 11.8 (13), 5.0 10.4 (25) 0.96, 0.38,
2008 (38) 160, 200 12, 12 (21.5) 0.167, 0.214, 0, (5.3), 8.9 0.96,
0.333 0.167 (17.7), 3.9 0.27,
(10.4), 0.86
4.9 (13)
All patients 61 36 51.5 (21.5) 51.4 (19.5) 0.115 0.111 NR 0.148 0.083 3.3 [0.69–16.2]‡ 7.0 (12.9) 10.4 (25) 0.48
Stein et al, Retrospective [8–140] 18§ 17§ 49.8 (26.9) 44.8 (20.6) 0.333 0.529 0.24 0.222 0.176 0.74 NR NR NR
2008 (52) comparative,
treatment
Nishijima et al, Retrospective 17.7 (6.2) 20 20 71 (IQR, 62–83) 81.5, (IQR, 0.350 0.350 1 0.200 0.050 0.15 NR NR NR
2010 (66)㛳 comparative, 70–87)
treatment
Brown et al, Retrospective NR 14 14 ¶ ¶ 0.500 0.286 0.22 0 0 1 NR NR NR
2010 (65)㛳 comparative,
treatment
Howes et al, Retrospective NR 36 36 62.5 (20.7) 62.4 (22.0) 0.528 0.222 0.014 0.138 0.056 0.28 NR NR NR
2009 (64)㛳 comparative,
treatment
Tawil et al, Retrospective NR 31 353 Age for treatment NR NR NR NR ** 3.1 [1.1–8.0]** NR NR NR NR
2008 (63)㛳 comparative, and usual care
treatment combined: 36
[11–90]

IQR ⫽ interquartile range; NR ⫽ not reported; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* Thromboembolic event rates were calculated by dividing the number of thromboembolic events by the sample size not the number of patients who had thromboembolic events. Therefore, the rates reported here may differ slightly
from those reported in each study. The tests of statistical significance presented are those reported by the individual studies and are not based on the thromboembolic event rates reported in this table.
† P values presented are those reported by the individual studies.
‡ Odds ratio and 95% CI.
§ Data are for patients with isolated traumatic brain injury only.
㛳 Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative analyses described in the Effectiveness Review Outcomes section and
the Evaluation of Harms section.
¶ Reported only for both groups combined: 50 y (SD, 21).
** Odds ratio and 95% CI for rFVIIa group vs. usual care group. Multivariate analyses demonstrate higher adjusted odds ratio of cerebral infarction with rFVIIa treatment.

19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 W-187

 Appendix Table 9. Mortality and Thromboembolic Events in Comparative Studies of rFVIIa Use in Liver Transplantation

Study, Year Study Design and rFVIIa Use Exact or Sample Size, n Mean Age (SD) [Range], y Mortality Rate Thromboembolic Event Rate*
(Reference) Mean (SD)
[Range] rFVIIa rFVIIa Group Usual rFVIIa Group Usual Care rFVIIa Group Usual P Value† rFVIIa Group Usual P Value†
Dose, mcg/kg Care Group Care Care
Group Group Group

Lodge et al, RCT, prophylaxis 180‡, 360‡ 63, 58 62 53.3 (11.2), 52.3 (11.5) 0.016§, 0.016§ NR 0.190㛳, 0.121㛳 0.097㛳 NS
2005 (111) 52.6 (9.2) 0.034§
Planinsic et al, RCT, prophylaxis 20, 40, 80 18, 24, 22 19 49.4 (13.4), 49.9 (11) 0.056§, 0.053§ NR 0.222, 0.083, 0.158 NR
2005 (40) 49.7 (10.1), 0.083§, 0.091
51.9 (8.8) 0.045§

W-188 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8

Pugliese et al, RCT, prophylaxis 40 10 10 NR NR 0 0 NR 0 0 NR
2007 (41)
Liu et al, RCT, prophylaxis [70–80] 14 14 51.9 [36–54] 47.5 [41–65] NR NR NR 0 NR NR
2009 (42)
Hendriks et al, Cohort, prophylaxis 80 6 12 43¶ [37–61] 48¶ [34–63] NR NR NR 0.167 0 NR
2001 (53)
De Gasperi et al, Retrospective comparative, [20–40] 6 6 45 (4) 47 (9) 0 0 NR NR NR NR
2005 (69)** prophylaxis
Kaliciński et al, Retrospective comparative, 52 [30–100] 28 61 13 (4) 11 (6) 0.071 0.082 NR 0 0.098 NR
2005 (67)** prophylaxis
Niemann et al, Retrospective comparative, 58 (18) 11 11 48 (15) 41 (17) 0 0.09 NR 0 0.09 NR
2006 (68)** prophylaxis

NR ⫽ not reported; NS ⫽ not significant; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* Thromboembolic event rates were calculated by dividing the number of thromboembolic events by the sample size not the number of patients who had thromboembolic events. Therefore, the rates reported here may differ slightly
from those reported in each study. The tests of statistical significance presented are those reported by the individual studies and are not based on the thromboembolic event rates reported in this table.
† The P values presented are those reported by the individual studies.
‡ Most patients in Lodge et al (111) received 3 doses of either 60- or 120-mcg/kg rFVIIa (range, 3– 6 doses).
§ Mortality rate is underestimated in Lodge et al (111) and Planinsic et al (40) because 2 additional deaths were not reported by trial group in each study.
㛳 Lodge et al (111) reports arterial thromboembolic events only.
¶ Median.
** Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes
section and the Evaluation of Harms section.

www.annals.org
 www.annals.org
Appendix Table 10. Indirect Outcomes in Comparative Studies of rFVIIa Use in Liver Transplantation

Study, Year Study Design and rFVIIa Exact or Sample Size, n Mean Age (SD) [IQR], y Mean RBC Transfusion (SD) [IQR], units OR Time (SD), min ICU Length of Stay (SD), d
(Reference) Use Mean (SD)
[Range] rFVIIa Group Usual rFVIIa Group Usual Care rFVIIa Group Usual Care Group P Value* rFVIIa Group Usual Care P Value* rFVIIa Usual Care P Value*
rFVIIa Dose, Care Group Group Group Group
mcg/kg Group

Lodge et al, RCT, prophylaxis 180†, 360† 62‡, 56‡ 61 53.3 (11.2), 52.3 (11.5) 7.0§, [0–76.5]㛳, 8.2§ [1.5–100.0]㛳 NS Same¶ Same¶ NS 3.5, 3.0 3.0 NS
2005 (111) 52.6 (9.2) 6.3§, [0–76.4]㛳
Planinsic et al, RCT, prophylaxis 20, 40, 80 18, 23, 22 19 49.4 (13.4), 49.9 (11.0) 10.0§, [3.0–18.0], 11.1§ [7.0–17.0] NS Same¶ Same¶ NS Same¶ Same¶ NS
2005 (40) 49.7 13.0§,
(10.1), [7.0–24.0],
51.9 (8.8) 10.0§, [3.2–21.0]
Pugliese et al, RCT, prophylaxis 40 10 10 – – 1.2 (0.53) 2.3 (0.44) 0.02 408 (56) 432 (48) NS 4.8 (1.3) 5.2 (1.2) NS
2007 (41)
Liu et al, RCT, prophylaxis [70–80] 14 14 51.9 [36–54] 47.5 [41–65] ** ** ** 268 (42) 485 (65) ⬍0.01 NR NR NR
2009 (42)
Hendriks et al, Cohort, prophylaxis 80 6 12 43§ [37–61]㛳 48§ [34–63]㛳 3.0§ [0–5]㛳 9.0§ [4–40]㛳 0.002 554 598 0.26 NR NR NR
2001 (53)
De Gasperi et al, Retrospective comparative, [20–40] 6 6 45 (4) 47 (9) 9 (4) 7 (2.5) NS NR NR NR NR NR NR
2005 (69)†† prophylaxis
Kaliciński et al, Retrospective comparative, 52 [30–100] 28 61 13 (4) 11 (6) ‡‡ ‡‡ ‡‡ 522 (110) 534 (118) 0.35 7.03 (5.94) 6.15 (5.55) 0.32
2005 (67)†† prophylaxis
Niemann et al, Retrospective comparative, 58 (18) 11 11 48 (15) 41 (17) 3.9 (2.6) 6.9 (2.3) 0.01 416 (104) 430 (84) 0.7 §§ §§ §§
2006 (68)†† prophylaxis

ICU ⫽ intensive care unit; IQR ⫽ interquartile range; NR ⫽ not reported; NS ⫽ not significant; OR ⫽ operating room; RBC ⫽ red blood cell; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* P values presented are those reported by the individual studies.
† Most patients in Lodge et al (111) received 3 doses of either 60- or 120-mcg/kg rFVIIa (range, 3– 6 doses).
‡ 1 patient in the 180-mcg/kg group and 2 patients in the 360-mcg/kg group were excluded from the analysis.
§ Median.
㛳 Range.
¶ Both studies stated, “No notable differences between study groups were found with respect to . . . operation duration,” (Lodge et al [111] and Planinsic et al [40]) or that “the mean number of days in the intensive care unit
was similar between studies” (Lodge et al [111]).
** Liu et al (42) does not report on RBC transfusion but does report blood loss: for rFVIIa, 2250 mL (SD, 350); for placebo, 6214 mL (SD, 750) (P ⬍ 0.01).
†† Studies did not meet inclusion criteria for detailed review in the comparative effectiveness analyses owing to poor quality but were included in the qualitative sensitivity analyses described in the Effectiveness Review Outcomes
section and the Evaluation of Harms section.
‡‡ Total intraoperative blood transfusion volume (not isolated to RBCs): for rFVIIa, 1980 mL; for usual care, 1527 mL (P ⫽ NS).
§§ Total hospital length of stay (not isolated to ICU length of stay): for rFVIIa, 11 days (SD, 7.3); for usual care, 7.9 days (SD, 2.7) (P ⫽ 0.2).

19 April 2011 Annals of Internal Medicine Volume 154 • Number 8 W-189

 Appendix Table 11. Mortality and Thromboembolic Events in RCTs on rFVIIa Use in Prostatectomy

Study, Year Study Design and Exact rFVIIa Sample Size, n Mean Age (SD), y Mortality Rate Thromboembolic Event Rate*
(Reference) rFVIIa Use Dose,
mcg/kg rFVIIa Group Usual Care rFVIIa Group Usual Care rFVIIa Group Usual Care P Value† rFVIIa Group Usual Care P Value†
Group Group Group Group
Friederich et al RCT, prophylaxis 20, 40 8, 16 12 61 (8.9), 63 (8.3) 0, 0 0 NR 0.125, 0 0 NR
2003 (112) 64 (8.5)

NR ⫽ not reported; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* Thromboembolic event rates were calculated by dividing the number of thromboembolic events by the sample size not the number of patients who had thromboembolic events. Therefore, the rates reported here may differ slightly
from those reported in each study. The tests of statistical significance presented are those reported by the individual studies and are not based on the thromboembolic event rates reported in this table.
† P values presented are those reported by the individual studies.

W-190 19 April 2011 Annals of Internal Medicine Volume 154 • Number 8

Appendix Table 12. Indirect Outcomes in RCTs on rFVIIa Use in Prostatectomy

Study, Year Study Design and Exact rFVIIa Sample Size, n Mean Age (SD), y RBC Transfusion (SD), units OR Time (SD), min
(Reference) rFVIIa Use Dose,
mcg/kg rFVIIa Group Usual Care rFVIIa Group Usual Care rFVIIa Group Usual Care P Value* rFVIIa Group Usual Care P Value*
Group Group Group Group
Friederich et al RCT, prophylaxis 20, 40 8, 16 12 61 (8.9), 63 (8.3) 0.6 (0.3), 1.5 (0.4) 0.057, 0.0003 126 (21), 120 (15) 180 (16) 0.034, 0.014
2003 (112) 64 (8.5) 0 (0)

OR ⫽ operating room; RBC ⫽ red blood cell; RCT ⫽ randomized, controlled trial; rFVIIa ⫽ recombinant factor VIIa.
* P values presented are those reported by the individual studies.

www.annals.org