- Email: [email protected]
Safety of Fibrinogen Concentrate and Cryoprecipitate in Cardiovascular Surgery: Multicenter Database Study
Accepted Manuscript
Safety of fibrinogen concentrate and cryoprecipitate in cardiovascular surgery: multi-center database study Takuma Maeda MD, MPH, PhD , Shigeki Miyata MD, PhD , Akihiko Usui MD, PhD , Kimitoshi Nishiwaki MD, PhD , Hitoshi Tanaka MD, PhD , Yutaka Okita MD, PhD , Nobuyuki Katori MD, PhD , Hideyuki Shimizu MD, PhD , Hiroaki Sasaki MD, PhD , Yoshihiko Ohnishi MD , Yuichi Ueda MD,PhD PII: DOI: Reference:
S1053-0770(18)30388-4 10.1053/j.jvca.2018.06.001 YJCAN 4749
To appear in:
Journal of Cardiothoracic and Vascular Anesthesia
Received date:
20 March 2018
Please cite this article as: Takuma Maeda MD, MPH, PhD , Shigeki Miyata MD, PhD , Akihiko Usui MD, PhD , Kimitoshi Nishiwaki MD, PhD , Hitoshi Tanaka MD, PhD , Yutaka Okita MD, PhD , Nobuyuki Katori MD, PhD , Hideyuki Shimizu MD, PhD , Hiroaki Sasaki MD, PhD , Yoshihiko Ohnishi MD , Yuichi Ueda MD,PhD , Safety of fibrinogen concentrate and cryoprecipitate in cardiovascular surgery: multi-center database study, Journal of Cardiothoracic and Vascular Anesthesia (2018), doi: 10.1053/j.jvca.2018.06.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Safety of fibrinogen concentrate and cryoprecipitate in cardiovascular surgery: multi-center database study Takuma Maeda, MD, MPH, PhD1,2,*[email protected], Shigeki Miyata, MD, PhD2, Akihiko Usui, MD, PhD3, Kimitoshi Nishiwaki, MD, PhD4, Hitoshi Tanaka, MD, PhD5,Yutaka Okita, MD, PhD5, Nobuyuki Katori, MD, PhD6, Hideyuki Shimizu, MD, PhD7, Hiroaki Sasaki, MD, PhD8,
aDepartment
of Anesthesiology, National Cerebral and Cardiovascular Center, Osaka, Japan
of Transfusion Medicine, National Cerebral and Cardiovascular Center, Osaka, Japan
cDepartment
of Thoracic Surgery, Nagoya University Hospital, Nagoya, Japan
dDepartment
of Anesthesiology, Nagoya University Hospital, Nagoya, Japan
eDepartment
of Cardiovascular Surgery, Kobe University, Kobe, Japan
AN US
bDivision
CR IP T
Yoshihiko Ohnishi, MD1 ,Yuichi Ueda, MD,PhD9
f
Department of Anesthesiology, Keio University, Tokyo, Japan
gDepartment hDepartment
of Cardiovascular Surgery, Keio University, Tokyo, Japan
of Cardiovascular Surgery, National Cerebral and Cardiovascular Center, Osaka, Japan
*Correspondence
of Nara Prefecture General Medical Center, Nara, Japan
M
iPresident
author: Department of Anesthesiology and Division of Transfusion Medicine,
Abstract
To
investigate
whether
administering
fibrinogen
concentrate
or
CE
Objectives:
PT
Fax: +81-6-6872-8175
ED
National Cerebral and, Cardiovascular Center, 5-7-1 Fujishirodai, Suita, Osaka 565-8565, Japan,
cryoprecipitate is associated with increased postoperative thromboembolic events and improved mortality in patients undergoing thoracic aortic surgery.
AC
Design: Multi-center retrospective cohort study using propensity-score analyses and multivariate logistic regression analysis to control for confounders. Setting: Four hospitals (one national cardiovascular center and three university hospitals).
Participants: Patients undergoing thoracic aortic surgery with cardiopulmonary bypass between January 2010 and October 2012 (n=1047). Interventions: We compared outcomes in patients treated with fibrinogen concentrate or cryoprecipitate (fibrinogen group) with those not receiving these products (no-fibrinogen group) based on propensity score matching. Multivariate logistic regression analysis 1
ACCEPTED MANUSCRIPT
was then performed to confirm the results. Measurements and Main Results: Among 1047 patients enrolled in this study, 247 patients received fibrinogen concentrate or cryoprecipitate. The median amount of administered fibrinogen was 3 g (interquartile range: 2–4 g). Eighty-seven patients were excluded from the propensity score matching because of missing data. Propensity score-matched analysis showed no significant difference in the incidence of
CR IP T
thromboembolic events or 30-day mortality rate between the groups. Multivariate analysis revealed that the fibrinogen group showed no significant difference in thromboembolic events (odds ratio, 1.22; 95% confidence interval, 0.76–1.95; P=0.408) or mortality rate (odds ratio, 0.44; 95% confidence interval, 0.18–1.12; P=0.081) compared with those in the no-fibrinogen group.
Conclusions: Administering fibrinogen concentrate or cryoprecipitate was associated
AN US
with neither thromboembolic events nor 30-day mortality in patients undergoing thoracic aortic surgery. Administering fibrinogen concentrate or cryoprecipitate are safe and do not appear to increase thromboembolic events and mortality in thoracic aortic surgery patients.
ED
cardiovascular surgery.
M
Keywords: fibrinogen; fibrinogen concentrate; cryoprecipitate; massive bleeding;
Complex cardiovascular surgery with cardiopulmonary bypass (CPB) can be associated
PT
with excessive bleeding1. Because several previous studies reported an association between bleeding and mortality2
3,
an optimal hemostatic strategy is desirable.
CE
Fibrinogen, which reaches critically low levels first among the clotting factors during
AC
major blood loss and volume replacement4, is a promising hemostatic target in cardiovascular surgery5 6. The Japanese Red Cross, which is the only organization in Japan to collect and supply blood for use in transfusion, does not provide cryoprecipitate, and fibrinogen concentrate is approved in Japan only for patients with congenital fibrinogen deficiency. Therefore, some hospitals produce cryoprecipitate in their own transfusion divisions, or they administer fibrinogen concentrate off-label.
2
ACCEPTED MANUSCRIPT
Theoretically, administering fibrinogen concentrate or cryoprecipitate raises plasma fibrinogen levels rapidly without volume overload, reversing coagulopathic bleeding by enhancing the strength of blood clot formation, resulting in less intraoperative bleeding However, recent results of randomized control trials studying this issue are
contradictory
9 10 11, 12,
likely because of their relatively small sample sizes and the
CR IP T
7 8.
difficulty involved in enrolling patients with severe bleeding into randomized control trials. An added concern is potential thromboembolic complications that worsen patients’
outcomes
because
fibrinogen
concentrate
and
cryoprecipitate
are
AN US
procoagulatory drugs. In fact, several studies have shown an association between elevated plasma fibrinogen and risk of arterial and venous thrombosis
13 14.
However,
there is only limited evidence regarding the safety of fibrinogen concentrate and
M
cryoprecipitate in patients undergoing cardiac surgery.
We hypothesized that concentrated fibrinogen products (fibrinogen concentrate and are
associated
with
ED
cryoprecipitate)
improved
mortality
without
increased
postoperative thromboembolic events. The goal of this multicenter study was to
PT
investigate our hypothesis using the database from the Japanese nationwide registry
CE
for cardiovascular surgery, which we combined with data on the administration of fibrinogen concentrate or cryoprecipitate retrieved from electronic medical records in
AC
each hospital, analyzing the data using propensity score matching and logistic regression analysis.
Materials and Methods
Patients and study design
3
ACCEPTED MANUSCRIPT
This retrospective, multicenter, cohort study involved four hospitals: National Cerebral and Cardiovascular Center (Osaka, Japan), Keio University Hospital (Tokyo, Japan), Kobe University Hospital (Hyogo, Japan), and Nagoya University Hospital (Aichi, Japan). This study was approved by the ethics committee of each institution and met
CR IP T
the guidelines of the Helsinki Declaration. Written informed consent was waived by the board because of the anonymous nature of the data. We enrolled patients aged ≥ 20 years undergoing thoracic vascular surgery with CPB between January 2010 and December 2012. Thoracic vascular surgery was divided into four groups: root surgery,
AN US
arch surgery, thoracic surgery, and combined surgery. Root surgery included mainly ascending artery replacement with or without additional surgeries such as valve replacement or coronary artery bypass grafting. Arch surgery included mainly total
M
arch replacement with or without additional surgeries such as valve replacement or coronary artery bypass grafting. Thoracic surgery included mainly descending thoracic
ED
aneurysm and thoracoabdominal aneurysm repairs. Combined surgery included mainly
PT
patients undergoing total arch replacement and thoracoabdominal aneurysm repair.
CE
Data collection
Data were collected from the hospitals’ electronic medical databases, which are
AC
uploaded to the Japan Adult Cardiovascular Database. This database consists of clinical information for all patients undergoing cardiovascular surgery in almost all hospitals in Japan, the details of which were described elsewhere15. The variables contained in the Japan Adult Cardiovascular Database are almost identical to those in the Society for Thoracic Surgeons' National Database, which collects detailed clinical information for patients undergoing cardiac surgery. However, the Japan Adult Cardiovascular
4
ACCEPTED MANUSCRIPT
Database lacks data on the administration of fibrinogen concentrate or cryoprecipitate; therefore, we retrieved this data from electronic medical records in each hospital and incorporated the data into the analyses.
CR IP T
Outcomes
The primary outcome was thromboembolic events (postoperative stroke, myocardial infarction, pulmonary embolism, and limb ischemia) during intensive care unit (ICU) stay. We defined thromboembolic events as: postoperative stroke, newly-emerged
AN US
paralysis of the central nerve system persisting more than 72 hours before discharge; postoperative myocardial infarction, newly-elevated cardiac biomarkers (creatine phosphokinase isoenzymes or creatine kinase MB isoenzyme greater than twice the
M
upper limit of the reference range, or elevated troponin); and postoperative pulmonary embolism diagnosed by lung perfusion scintigraphy or angiography; and limb ischemia
ED
and any complication caused by limb ischemia. The secondary outcome was 30-day
CE
PT
mortality.
Management of anticoagulation, and cardiopulmonary bypass
AC
This was an observational study, and anticoagulation and cardiopulmonary bypass management were neither controlled nor standardized. Usually, we administered 300– 400 U/kg of unfractionated heparin as a priming dose for anticoagulation for CPB to achieve an activated clotting time > 400 seconds. After CPB discontinuation, we used protamine to neutralize heparin. Postoperatively, patients were transferred to the ICU for monitoring and mechanical
5
ACCEPTED MANUSCRIPT
ventilation; patients were extubated when standard criteria were met. Usually, dynamic mechanical thromboprophylaxis (intermittent compression devices or pneumatic compression devices) or graduated compression stockings were used
CR IP T
postoperatively if there were no contraindications.
Transfusion practices
This was an observational study, and transfusion practices were neither controlled nor standardized. We usually administer red blood cells with a target hemoglobin level of 9–
AN US
10 g/dL, and this policy did not change through the study period. The administration of fresh-frozen plasma (FFP) and platelet concentrate were at the discretion of attending anesthesiologists
and
cardiac surgeons.
Usually,
FFP
was
administered
for
M
international normalized ratios > 1.5, and platelet concentrate was administered for platelet counts < 100 000 platelets/µL. In the National Cerebral and Cardiovascular
ED
Center, cryoprecipitate was transfused if fibrinogen was < 150 mg/dL. Usually, one pool of cryoprecipitate was prepared from 1440 ml of FFP, and the pool contained
PT
approximately 2 g of fibrinogen. In Nagoya University Hospital and Kobe University
CE
Hospital, fibrinogen concentrate (Fibrinogen HT; Japan Blood Products Organization, Tokyo, Japan) was administered if fibrinogen was < 150 mg/dL at weaning from CPB or
AC
when the attending surgeon believed that fibrinogen concentrate was required because of surgical bleeding. In Keio University Hospital, neither fibrinogen concentrate nor cryoprecipitate was used intra- or postoperatively; only FFP was administered, targeting a fibrinogen concentration of 120–150 mg/dl based on surgical hemostasis. As an antifibrinolytic agent, tranexamic acid was used at the anesthesiologist’s discretion.
6
ACCEPTED MANUSCRIPT
Statistics First, we eliminated patients with missing characteristics data (Fig 1). Then, we performed propensity score (PS) matching to balance patients’ backgrounds between the fibrinogen administration group (patients receiving fibrinogen concentrate or
CR IP T
cryoprecipitate) and no-fibrinogen group (patients not receiving fibrinogen concentrate or cryoprecipitate). PS was calculated to predict the probability of receiving fibrinogen concentrate or cryoprecipitate based on the predictors in Table 1 using a logistic regression model. We performed one-to-one patient matching between fibrinogen and
AN US
no-fibrinogen groups with the closest estimated PS within one caliper (≤ 0.2 of the pooled standard deviation of estimated logits) using the nearest-neighbor method. We estimated the balance in baseline variables using standardized differences, where >
M
10% was considered imbalanced16. We compared 30-day mortality and perioperative thromboembolic events between the groups using the 2 test.
ED
Second, we performed multivariate logistic regression analysis to confirm the results. To calculate odds ratios (OR) and 95% confidence intervals (95% CI) for 30-day mortality
PT
and thromboembolic events during ICU stay, variables with a P-value < .2 with
CE
univariate logistic regression analysis were selected for the multivariate logistic regression model with a forced entry of use of fibrinogen concentrate or cryoprecipitate.
AC
Regarding “age”, we calculated odds per 10 years and, as for “duration of operation” and “duration of CPB”, we calculated odds per 1-hour increase, for easier understanding, clinically. Multiple imputation was used for the multivariate logistic regression analyses to account for variables with missing data (n=1047). We performed all statistical analyses using IBM SPSS, version 22 (IBM Corp., Armonk, NY).
7
ACCEPTED MANUSCRIPT
Results
CR IP T
We enrolled 1047 patients during the study period. Postoperative outcome parameters in the four hospitals are shown in the supplementary table (Supplementary Table). To perform PS matching, we excluded 87 patients because of missing characteristics data, and analyzed data for 960 study patients. Our PS matching created 191 pairs of
AN US
patients with and without fibrinogen concentrate or cryoprecipitate administration (Fig. 1). The median amount of administered fibrinogen was 3 g (interquartile range: 2–4 g) in the fibrinogen group, which was calculated by assuming that one pool of
M
cryoprecipitate contained 2 g of fibrinogen.
Table 1 shows the baseline characteristics of the unmatched and PS-matched groups.
ED
After PS matching, standardized differences for the measured covariates, except degree of urgency, type of surgery, and hospitals, were < 0.1, suggesting that groups were
PT
generally well-balanced. Table 2 shows the proportion of 30-day mortality and the
CE
incidence of thromboembolic events between the PS-matched groups. We found no significant difference in the incidence of both outcomes between the groups. Even with
AC
1:2 and 1:3 matching, we found no significant difference in the incidence of both outcomes between the groups (data not shown). Next, to confirm these results, we performed multivariate logistic regression analysis. We identified several variables that differed between patients who experienced perioperative thromboembolic events (N=105) and those who did not (N=942) (Table 3). Multivariate analysis revealed that carotid vessel disease (OR, 2.25; 95% CI, 1.24–4.07;
8
ACCEPTED MANUSCRIPT
P=0.008) and duration of operation (per 1-hour increase: OR, 1.17; 95% CI, 1.05–1.29; P=0.003) were significantly associated with thromboembolic events. However, administering fibrinogen concentrate or cryoprecipitate was not associated with thromboembolic events. Similarly, age (per 10-year increase: OR, 1.43; 95% CI, 1.03–
CR IP T
1.97; P=0.031), preoperative resuscitation (OR, 14.07; 95% CI, 3.77–52.5; P<0.001), and duration of operation (per 1-hour increase: OR, 1.20; 95% CI, 1.02–1.40; P=0.028) were independent risk factors for 30-day mortality (Thirty-five of 1047 patients died). However, administering fibrinogen concentrate or cryoprecipitate was not associated
M
AN US
with 30-day mortality (OR, 0.44; 95% CI, 0.18–1.12; P=0.081) (Table 4).
ED
Discussion
This study investigated whether administering fibrinogen concentrate or to
treat
PT
cryoprecipitate
bleeding
is
associated
with
increased
postoperative
CE
thromboembolic events and/or improved mortality in patients undergoing thoracic vascular surgery with CPB. In this multicenter retrospective study using national
AC
registry data, PS matching revealed that administering fibrinogen concentrate or cryoprecipitate was associated with neither thromboembolic events nor 30-day mortality, which was confirmed by multivariate logistic regression models with a multiple imputation method for missing data. Previous randomized controlled trials have attempted to confirm the efficacy, but not the safety, of fibrinogen concentrate. 9 10 11, 12 These studies, given a low event rate of
9
ACCEPTED MANUSCRIPT
adverse events, are rarely powered to address safety issues. For example, death was reported in only 2/60 patients in the fibrinogen group, and none occurred in the control group in a recent randomized controlled trial12. Another randomized controlled trial reported thromboembolic events in 6/78 patients (7.7%) in the fibrinogen group and
CR IP T
10/74 patients (13.5%) in the placebo group11. A recent retrospective single-center cohort study suggested that perioperative administration of fibrinogen concentrate to bleeding patients undergoing cardiac surgery is not associated with increased incidence of major adverse cardiac and thromboembolic events and mortality17, which is consistent with
AN US
our findings. Contrasting with the previous study, which investigated patients undergoing cardiac surgery (with the exception of patients undergoing surgery of the ascending aorta17), we investigated patients undergoing thoracic aortic surgery, which
M
can be a more complex surgery with a higher risk of massive bleeding. We also enrolled patients from four different institutions, which potentially contributed to external
ED
validity18.
The ideal clinical situation in which to use fibrinogen concentrate or
PT
cryoprecipitate remains unknown. Theoretically, fibrinogen is the first coagulation
CE
factor to reach critically low levels during major blood loss and volume replacement, functioning both as a source of fibrin and as a mediator of platelet aggregation 4.
AC
Therefore, it is reasonable to hypothesize that rapid increases in fibrinogen levels would be more beneficial in patients undergoing complex surgery with a high risk of massive bleeding. To test this hypothesis, we investigated patients who underwent thoracic aortic surgery. Fibrinogen concentrate and cryoprecipitate were not used at all in one hospital, which may have worked in favor of PS matching because neither fibrinogen concentrate nor cryoprecipitate was used, even in severe cases. Our results confirmed
10
ACCEPTED MANUSCRIPT
the safety of fibrinogen concentrate or cryoprecipitate in our chosen patient population. Multivariate logistic regression analysis revealed that the OR of fibrinogen or cryoprecipitate use for 30-day mortality missed statistical significance (OR, 0.44; 95% CI, 0.18–1.12; P=0.081). In our additional logistic regression analysis dividing
CR IP T
fibrinogen concentrate and cryoprecipitate as explanatory variables, we found that fibrinogen concentrate may effectively reduce mortality (OR, 0.20; 95% CI, 0.05–0.84). Further studies with a higher study power may clarify the effect of using fibrinogen concentrate or cryoprecipitate.
AN US
We acknowledge several limitations of this study. First, because of the observational nature of this study, inference of causality is limited. Also, patients may have left the study, and misdiagnosis of rare adverse events is possible. For example,
M
spinal cord ischemia might present as stroke and be misidentified as a thrombotic complication. It is also possible that some events could have been missed because of the
ED
absence of standardized protocols. Second, although this study was based on multicenter data, the data were collected retrospectively. Despite using PS matching to
PT
reduce bias related to confounding, we cannot eliminate the possibility that there may
CE
be unobserved confounders such as differences in perioperative management between hospitals, practices, other blood product management systems, and the high likelihood
AC
that bleeding patients receive fibrinogen. Our PS matching results are generalizable only among patients in the PS range included in the paired analysis, and these results may not be applicable to patients with scores outside this range. However, we conducted multivariate logistic regression analysis using a multiple imputation method for missing data, and confirmed the robustness of the results. Third, our database lacked data on antifibrinolytic agents such as tranexamic acid used intraoperatively and
11
ACCEPTED MANUSCRIPT
anticoagulation therapy with antiplatelet or anticoagulation agents preoperatively, which may also have affected our results. Fourth, we lack laboratory data for perioperative fibrinogen levels or viscoelastic tests, which may be useful in guiding appropriate use of fibrinogen products, and this may have affected our results. Fifth, we
CR IP T
experienced 105 thrombotic events and 35 deaths. Considering this small number of deaths, the logistic regression analysis may have been affected by overfitting.
In conclusion, this study revealed that treatment with fibrinogen concentrate or cryoprecipitate was associated with neither thromboembolic events nor 30-day
AN US
mortality in patients undergoing thoracic aortic surgery.
Acknowledgments
M
This research was supported in part by the Research Grant for Regulatory Science of
ED
Pharmaceuticals and Medical Devices from Japan Agency for Medical Research and Development
PT
(AMED) and a grant-in-aid from the Takeda Science Foundation. We thank Jane Charbonneau,
CE
DVM, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
AC
Source of support:
This research was supported in part by the Research Grant on Regulatory Science of
Pharmaceuticals and Medical Devices from Japan Agency for Medical Research and development,
AMED, and a grant-in-aid from the Takeda Science Foundation.
12
ACCEPTED MANUSCRIPT
Conflicts of interest:
Yuichi Ueda received speaker honoraria and consulting fee from CSL Behring K.K. (Tokyo, Japan).
Shigeki Miyata received research support and speaker honoraria from Daiichi Sankyo Co., Ltd.
CR IP T
(Tokyo,Japan), Mitsubishi Tanabe Pharma Corporation (Osaka, Japan), and CSL Behring K.K.
(Tokyo, Japan). None of the other authors declare any conflicts of interest.
1.
AN US
References
Sniecinski RM, Levy JH: Bleeding and management of coagulopathy. J
2.
M
Thorac Cardiovasc Surg 142:662-667,2011
Murphy GJ, Reeves BC, Rogers CA, et al.: Increased mortality,
ED
postoperative morbidity, and cost after red blood cell transfusion in patients having cardiac surgery. Circulation 116:2544-2552,2007 Koch CG, Li L, Duncan AI, et al.: Morbidity and mortality risk associated
PT
3.
CE
with red blood cell and blood-component transfusion in isolated coronary artery bypass grafting. Crit Care Med 34:1608-1616,2006 Hiippala ST, Myllyla GJ, Vahtera EM: Hemostatic factors and replacement
AC
4.
of major blood loss with plasma-poor red cell concentrates. Anesth Analg 81:360-365,1995 5.
Karkouti K, Callum J, Crowther MA, et al.: The relationship between
fibrinogen levels after cardiopulmonary bypass and large volume red cell transfusion
in
cardiac
surgery:
an 13
observational
study.
Anesth
Analg
ACCEPTED MANUSCRIPT
117:14-22,2013 6.
Collins PW, Solomon C, Sutor K, et al.: Theoretical modelling of fibrinogen
supplementation
with
therapeutic
plasma,
cryoprecipitate,
or
fibrinogen
concentrate. Br J Anaesth 113:585-595,2014 Bolliger D, Szlam F, Molinaro RJ, et al.: Finding the optimal concentration
CR IP T
7.
range for fibrinogen replacement after severe haemodilution: an in vitro model. Br J Anaesth 102:793-799,2009 8.
Solomon C, Pichlmaier U, Schoechl H, et al.: Recovery of fibrinogen after
AN US
administration of fibrinogen concentrate to patients with severe bleeding after cardiopulmonary bypass surgery. Br J Anaesth 104:555-562,2010 9.
Rahe-Meyer N, Solomon C, Hanke A, et al.: Effects of fibrinogen
M
concentrate as first-line therapy during major aortic replacement surgery: a randomized, placebo-controlled trial. Anesthesiology 118:40-50,2013 Ranucci
M,
Baryshnikova
ED
10.
E,
Crapelli
GB,
et
al.:
Randomized,
double-blinded, placebo-controlled trial of fibrinogen concentrate supplementation
Rahe-Meyer N, Levy JH, Mazer CD, et al.: Randomized evaluation of
CE
11.
PT
after complex cardiac surgery. J Am Heart Assoc 4:e002066,2015
fibrinogen vs placebo in complex cardiovascular surgery (REPLACE): a double-blind
AC
phase III study of haemostatic therapy. Br J Anaesth 117:41-51,2016 12.
Bilecen S, de Groot JA, Kalkman CJ, et al.: Effect of Fibrinogen
Concentrate on Intraoperative Blood Loss Among Patients With Intraoperative Bleeding During High-Risk Cardiac Surgery: A Randomized Clinical Trial. JAMA 317:738-747,2017 13.
Wilhelmsen L, Svardsudd K, Korsan-Bengtsen K, et al.: Fibrinogen as a 14
ACCEPTED MANUSCRIPT
risk factor for stroke and myocardial infarction. N Engl J Med 311:501-505,1984 14.
van Hylckama Vlieg A, Rosendaal FR: High levels of fibrinogen are
associated with the risk of deep venous thrombosis mainly in the elderly. J Thromb Haemost 1:2677-2678,2003 Okita Y, Miyata H, Motomura N, et al.: A study of brain protection during
CR IP T
15.
total arch replacement comparing antegrade cerebral perfusion versus hypothermic circulatory arrest, with or without retrograde cerebral perfusion: analysis based on the Japan Adult Cardiovascular Surgery Database. J Thorac Cardiovasc Surg
16.
AN US
149:S65-73,2015
Austin PC: The use of propensity score methods with survival or
time-to-event outcomes: reporting measures of effect similar to those used in
17.
M
randomized experiments. Stat Med 33:1242-1258,2014
Fassl J, Lurati Buse G, Filipovic M, et al.: Perioperative administration of
ED
fibrinogen does not increase adverse cardiac and thromboembolic events after cardiac surgery. Br J Anaesth 114:225-234,2015 Bellomo R, Warrillow SJ, Reade MC: Why we should be wary of
PT
18.
AC
CE
single-center trials. Crit Care Med 37:3114-3119,2009
15
AC
CE
PT
ED
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
Fig. 1. Flow diagram of patient enrollment
Table 1. Baseline Patient Characteristics in the Unmatched and Propensity-Matched 16
ACCEPTED MANUSCRIPT
Groups Unmatched Groups Fib/Cryo
No
Standardi
Fib/Cryo
No
Standardi
administr
Fib/Cryo
zed
administr
Fib/Cryo
zed
ation
administr
Differenc
ation
administr
Differenc
(n=247)
ation
es, %
(n=191)
ation
es, %
(n=713)
(n=191)
CR IP T
variable
Matched Groups
Age (year)
67 (58-75)
69 (59-77)
-1.8
67 (59-76)
69 (60-77)
-2.7
Gender (%
163 (66.0)
487 (68.3)
-4.9
128 (67.0)
126 (66.0)
2.2
Height
163
164
-7.9
164
162
6.7
(cm)
(156-169)
(157-171)
Weight (kg)
60.3
61.6
(51-69)
(53-70)
124 (50.2)
345 (48.4)
64 (25.9)
185 (25.9)
177 (71.7)
535 (75.0)
27 (10.9)
males)
91 (47.6)
86 (45.0)
4.8
-0.1
46 (24.1)
51 (26.7)
-6.0
-7.6
136 (71.2)
129 (67.5)
8.0
77 (10.8)
0.4
19 (9.9)
22 (11.5)
-5.1
27 (10.9)
62 (8.7)
7.5
21 (11.0)
23 (12.0)
-3.3
2 (0.8)
5 (0.7)
1.3
4 (2.1)
2 (1.0)
8.4
28 (11.3)
61 (8.6)
9.3
24 (12.6)
23 (12.0)
1.6
8 (3.2)
14 (2.0)
8.0
6 (3.1)
5 (2.6)
3.1
19 (7.7)
62 (8.7)
-3.7
12 (6.3)
17 (8.9)
9.8
failure (%) Liver
M
ED
CE
dysfunctio
PT
(%) Renal
2.9
3.6
on (%) Diabetes
58.9
(51.5-67.0)
emia (%) Hypertensi
60.4
(52.0-69.5)
(%) Hyperlipid
(156-170)
AN US
Smoking
-10.5
(157-170)
n (%)
AC
Cerebrovas cular
accident (%)
Preoperati ve resuscitati on (%) COPD
17
ACCEPTED MANUSCRIPT
moderate to
severe 30 (12.1)
76 (10.7)
4.7
21 (11.0)
16 (8.4)
8.9
13 (5.3)
26 (3.6)
7.8
7 (3.7)
5 (2.6)
6.0
11 (4.4)
8 (1.1)
20.3
6 (3.1)
4 (2.1)
6.6
66 (26.7)
109 (15.3)
28.3
46 (24.1)
40 (20.9)
7.5
CCS (II-IV)
12 (4.9)
21 (2.9)
9.9
8 (4.2)
8 (4.2)
0
NYHA
182 (73.4)
437 (61.3)
26.7
137 (71.7)
133 (69.6)
4.6
good
182 (73.7)
M
(%) Peripheral
518 (72.7)
2.3
141 (73.8)
143 (74.9)
-2.4
moderate
60 (24.3)
185 (25.9)
-3.8
46 (24.1)
45 (23.6)
1.2
poor
5 (2.0)
10 (1.4)
4.8
4 (2.1)
3 (1.6)
3.9
Duration of
497
412
54.6
473
434
4.0
operation
(390-630)
(349-502)
(381-573)
(372-547)
Duration of
237
203
225
218
CPB (min)
(189-307)
(161-245)
(183-287)
(182-263)
Preoperati
9 (3.6)
25 (3.5)
0.7
7 (3.7)
7 (3.7)
0
elective
158 (64.0)
546 (76.6)
-27.9
127 (66.5)
112 (58.6)
16.3
urgent
20 (8.1)
41 (5.8)
9.3
10 (5.2)
20 (10.5)
-19.6
emergent
66 (26.7)
123 (17.3)
23.0
52 (27.2)
57 (29.8)
-5.7
salvage
3 (1.2)
3 (0.4)
8.8
2 (1.0)
2 (1.0)
0
vascular Old myocardial infarction (%) Congestive failure (%) Reoperatio n (%)
(II-IV)
AC
PT
CE
(min)
ED
LV function
ve
AN US
heart
CR IP T
disease (%)
54.9
9.4
shock
(%)
Urgency
18
ACCEPTED MANUSCRIPT
Arrhythmi
24 (9.7)
65 (9.1)
2.1
18 (9.4)
17 (8.9)
1.8
Lowest
24.5(22.1-
25.0
-18.2
24.5
24.0
-2.9
core
26.4)
(22.0-28.0)
(22.0-26.6)
(21.2-26.7)
a (%)
temperatur Type of cardiac surgery
CR IP T
e (˚C) Root
50 (20.2)
161 (22.6)
-5.7
38 (19.9)
34 (17.8)
5.4
arch
134 (54.3)
410 (57.5)
-6.6
102 (53.4)
118 (61.8)
-17.0
thoracic
51 (20.6)
120 (16.8)
9.8
43 (22.5)
29 (15.2)
18.8
combined
12 (4.9)
22 (3.1)
9.1
8 (4.2)
10 (5.2)
-4.9
NCVC
103 (41.7)
336 (47.1)
-10.9
99 (51.8)
65 (34.0)
36.6
Keio
0 (0)
198 (27.8)
-87.7
0 (0)
37 (19.4)
-69.3
30 (12.1)
135 (18.9)
-18.8
25 (13.1)
60 (31.4)
-45.2
114 (46.2)
44 (6.2)
67 (35.1)
29 (15.2)
-47.1
University Kobe University Nagoya
102.1
ED
Hospital
M
Hospital
AN US
Hospital
University Hospital
PT
Numerical data are presented as medians (1st quartile–3rd quartile). Categorical data are presented as number (%). Abbreviations: Fib, fibrinogen concentrate; Cryo, cryoprecipitate; COPD, chronic
CE
obstructive pulmonary disease; CCS, Canadian Cardiovascular Society functional classification; NYHA, New York Heart Association; NCVC, National Cerebral
AC
Cardiovascular Center
Table 2. Postoperative outcome parameters in the propensity-score-matched groups 19
ACCEPTED MANUSCRIPT
characteristic Thrombosis
Fibrinogen/cryoprecipitate
No Fibrinogen/cryoprecipitate
P
group (n=191)
group(n=191)
27 (14.1)
28 (14.7)
0.884
6 (3.1)
10 (5.2)
0.307
(%) 30-day mortality (%)
CR IP T
Categorical data are presented as number (%).
Table 3. Multivariate logistic regression analysis of patients' thrombosis events in the intensive care units Univariate analysis
Multivariate analysis
OR
95% CI
Fibrinogen/
1.65
1.08-2.52
0.021
95% CI
P
1.22
0.76-1.95
0.408
1.10
0.95-1.27
0.192
1.12
0.95-1.32
0.188
Gender (male)
1.55
0.97-2.46
0.064
1.22
0.71-2.07
0.474
Smoking (missing 1)
1.59
1.06-2.40
0.026
1.26
0.78-2.03
0.342
DM
1.66
0.95-2.91
0.075
1.37
0.74-2.53
0.323
Carotid vessel disease
2.29
1.31-4.00
0.004
2.25
1.24-4.07
0.008
1.76
0.94-3.30
0.078
1.37
0.70-2.70
0.362
1.51
0.88-2.59
0.139
1.20
0.66-2.18
0.560
1.75
0.76-4.02
0.190
1.03
0.40-2.68
0.955
1.39
0.86-2.24
0.178
1.02
0.59-1.78
0.943
1.14
1.07-1.22
<0.001
1.17
1.05-1.29
0.003
1.19
1.06-1.35
0.004
0.99
0.78-1.14
0.559
cryoprecipitate use Age (per 10 years
PT
Peripheral vascular
ED
severe
M
increase)
COPD moderate to
P
OR
AN US
covariates
disease (missing 51)
CE
Old myocardial
infarction (missing 2)
AC
Reoperation
Duration of operation (per 1 hour increase) Duration of CPB (per 1 hour increase) (missing 77)
Type of cardiac surgery (missing 31) Root
1
1
(reference)
(reference) 20
ACCEPTED MANUSCRIPT
Arch
1.00
0.60-1.67
0.999
0.76
0.44-1.30
0.307
Thoracic
0.86
0.44-1.68
0.660
0.57
0.27-1.21
0.142
Combined
2.80
1.14-6.90
0.025
1.75
0.68-4.54
0.249
Abbreviations: OR, odds ratio; CI, confidence interval; DM, diabetes mellitus; COPD, chronic obstructive pulmonary disease; CPB, cardiopulmonary bypass; missing (n),
AC
CE
PT
ED
M
AN US
CR IP T
missing data for the number of patients (n)
21
ACCEPTED MANUSCRIPT
Table 4. Multivariate logistic regression analysis of 30-day mortality Multivariate analysis
OR
95% CI
P
OR
95% CI
P
Fibrinogen/
1.11
0.52-2.33
0.794
0.44
0.18-1.12
0.081
1.22
0.94-1.58
0.133
1.42
1.03-1.95
0.03
Renal failure
2.09
0.84-0.18
0.109
2.02
0.71-3.35
0.16
Carotid vessel
2.59
1.10-6.11
0.029
2.08
0.74-5.45
0.17
16.6
6.3-43.9
<0.001
13.0
3.59-47.3
<0.001
2.51
1.01-6.24
0.047
2.54
0.90-7.16
0.08
1.23
<0.001
1.20
1.02-1.43
0.02
<0.001
1.23
0.91-1.65
0.17
CR IP T
covariates
M
Univariate analysis
cryoprecipitate use Age (per 10 years
AN US
increase)
disease Resuscitation (missing 1) COPD moderate to severe Duration of
ED
operation (per 1 hour increase) (per 1 hour
1.38
1.17-1.63
PT
Duration of CPB
1.13-1.34
77)
CE
increase) (missing Urgency
AC
elective
1
1
(reference)
(reference)
Urgent
1.05
0.24-4.59
0.946
0.67
0.11-4.12
0.66
Emergent
2.16
1.02-4.59
0.045
1.25
0.77-2.05
0.65
Salvage
18.7
3.23-108.0
<0.001
1.59
0.17-14.7
0.68
0.16-0.96
0.04
Type of cardiac surgery (missing 31) Root Arch
1
1
(reference)
(reference)
0.51
0.22-1.14 22
0.102
0.39
ACCEPTED MANUSCRIPT
Thoracic
0.84
0.36-2.31
0.839
0.81
0.26-2.56
0.72
Combined
1.26
0.27-5.93
0.774
0.77
0.33-1.81
0.76
Abbreviations: CI, confidence interval; P, p-value; COPD, chronic obstructive pulmonary disease; CPB, cardiopulmonary bypass; missing (n), missing data for the number of
AC
CE
PT
ED
M
AN US
CR IP T
patients (n)
23
Safety of fibrinogen concentrate and cryoprecipitate in cardiovascular surgery: multi-center database study Takuma Maeda MD, MPH, PhD , Shigeki Miyata MD, PhD , Akihiko Usui MD, PhD , Kimitoshi Nishiwaki MD, PhD , Hitoshi Tanaka MD, PhD , Yutaka Okita MD, PhD , Nobuyuki Katori MD, PhD , Hideyuki Shimizu MD, PhD , Hiroaki Sasaki MD, PhD , Yoshihiko Ohnishi MD , Yuichi Ueda MD,PhD PII: DOI: Reference:
S1053-0770(18)30388-4 10.1053/j.jvca.2018.06.001 YJCAN 4749
To appear in:
Journal of Cardiothoracic and Vascular Anesthesia
Received date:
20 March 2018
Please cite this article as: Takuma Maeda MD, MPH, PhD , Shigeki Miyata MD, PhD , Akihiko Usui MD, PhD , Kimitoshi Nishiwaki MD, PhD , Hitoshi Tanaka MD, PhD , Yutaka Okita MD, PhD , Nobuyuki Katori MD, PhD , Hideyuki Shimizu MD, PhD , Hiroaki Sasaki MD, PhD , Yoshihiko Ohnishi MD , Yuichi Ueda MD,PhD , Safety of fibrinogen concentrate and cryoprecipitate in cardiovascular surgery: multi-center database study, Journal of Cardiothoracic and Vascular Anesthesia (2018), doi: 10.1053/j.jvca.2018.06.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Safety of fibrinogen concentrate and cryoprecipitate in cardiovascular surgery: multi-center database study Takuma Maeda, MD, MPH, PhD1,2,*[email protected], Shigeki Miyata, MD, PhD2, Akihiko Usui, MD, PhD3, Kimitoshi Nishiwaki, MD, PhD4, Hitoshi Tanaka, MD, PhD5,Yutaka Okita, MD, PhD5, Nobuyuki Katori, MD, PhD6, Hideyuki Shimizu, MD, PhD7, Hiroaki Sasaki, MD, PhD8,
aDepartment
of Anesthesiology, National Cerebral and Cardiovascular Center, Osaka, Japan
of Transfusion Medicine, National Cerebral and Cardiovascular Center, Osaka, Japan
cDepartment
of Thoracic Surgery, Nagoya University Hospital, Nagoya, Japan
dDepartment
of Anesthesiology, Nagoya University Hospital, Nagoya, Japan
eDepartment
of Cardiovascular Surgery, Kobe University, Kobe, Japan
AN US
bDivision
CR IP T
Yoshihiko Ohnishi, MD1 ,Yuichi Ueda, MD,PhD9
f
Department of Anesthesiology, Keio University, Tokyo, Japan
gDepartment hDepartment
of Cardiovascular Surgery, Keio University, Tokyo, Japan
of Cardiovascular Surgery, National Cerebral and Cardiovascular Center, Osaka, Japan
*Correspondence
of Nara Prefecture General Medical Center, Nara, Japan
M
iPresident
author: Department of Anesthesiology and Division of Transfusion Medicine,
Abstract
To
investigate
whether
administering
fibrinogen
concentrate
or
CE
Objectives:
PT
Fax: +81-6-6872-8175
ED
National Cerebral and, Cardiovascular Center, 5-7-1 Fujishirodai, Suita, Osaka 565-8565, Japan,
cryoprecipitate is associated with increased postoperative thromboembolic events and improved mortality in patients undergoing thoracic aortic surgery.
AC
Design: Multi-center retrospective cohort study using propensity-score analyses and multivariate logistic regression analysis to control for confounders. Setting: Four hospitals (one national cardiovascular center and three university hospitals).
Participants: Patients undergoing thoracic aortic surgery with cardiopulmonary bypass between January 2010 and October 2012 (n=1047). Interventions: We compared outcomes in patients treated with fibrinogen concentrate or cryoprecipitate (fibrinogen group) with those not receiving these products (no-fibrinogen group) based on propensity score matching. Multivariate logistic regression analysis 1
ACCEPTED MANUSCRIPT
was then performed to confirm the results. Measurements and Main Results: Among 1047 patients enrolled in this study, 247 patients received fibrinogen concentrate or cryoprecipitate. The median amount of administered fibrinogen was 3 g (interquartile range: 2–4 g). Eighty-seven patients were excluded from the propensity score matching because of missing data. Propensity score-matched analysis showed no significant difference in the incidence of
CR IP T
thromboembolic events or 30-day mortality rate between the groups. Multivariate analysis revealed that the fibrinogen group showed no significant difference in thromboembolic events (odds ratio, 1.22; 95% confidence interval, 0.76–1.95; P=0.408) or mortality rate (odds ratio, 0.44; 95% confidence interval, 0.18–1.12; P=0.081) compared with those in the no-fibrinogen group.
Conclusions: Administering fibrinogen concentrate or cryoprecipitate was associated
AN US
with neither thromboembolic events nor 30-day mortality in patients undergoing thoracic aortic surgery. Administering fibrinogen concentrate or cryoprecipitate are safe and do not appear to increase thromboembolic events and mortality in thoracic aortic surgery patients.
ED
cardiovascular surgery.
M
Keywords: fibrinogen; fibrinogen concentrate; cryoprecipitate; massive bleeding;
Complex cardiovascular surgery with cardiopulmonary bypass (CPB) can be associated
PT
with excessive bleeding1. Because several previous studies reported an association between bleeding and mortality2
3,
an optimal hemostatic strategy is desirable.
CE
Fibrinogen, which reaches critically low levels first among the clotting factors during
AC
major blood loss and volume replacement4, is a promising hemostatic target in cardiovascular surgery5 6. The Japanese Red Cross, which is the only organization in Japan to collect and supply blood for use in transfusion, does not provide cryoprecipitate, and fibrinogen concentrate is approved in Japan only for patients with congenital fibrinogen deficiency. Therefore, some hospitals produce cryoprecipitate in their own transfusion divisions, or they administer fibrinogen concentrate off-label.
2
ACCEPTED MANUSCRIPT
Theoretically, administering fibrinogen concentrate or cryoprecipitate raises plasma fibrinogen levels rapidly without volume overload, reversing coagulopathic bleeding by enhancing the strength of blood clot formation, resulting in less intraoperative bleeding However, recent results of randomized control trials studying this issue are
contradictory
9 10 11, 12,
likely because of their relatively small sample sizes and the
CR IP T
7 8.
difficulty involved in enrolling patients with severe bleeding into randomized control trials. An added concern is potential thromboembolic complications that worsen patients’
outcomes
because
fibrinogen
concentrate
and
cryoprecipitate
are
AN US
procoagulatory drugs. In fact, several studies have shown an association between elevated plasma fibrinogen and risk of arterial and venous thrombosis
13 14.
However,
there is only limited evidence regarding the safety of fibrinogen concentrate and
M
cryoprecipitate in patients undergoing cardiac surgery.
We hypothesized that concentrated fibrinogen products (fibrinogen concentrate and are
associated
with
ED
cryoprecipitate)
improved
mortality
without
increased
postoperative thromboembolic events. The goal of this multicenter study was to
PT
investigate our hypothesis using the database from the Japanese nationwide registry
CE
for cardiovascular surgery, which we combined with data on the administration of fibrinogen concentrate or cryoprecipitate retrieved from electronic medical records in
AC
each hospital, analyzing the data using propensity score matching and logistic regression analysis.
Materials and Methods
Patients and study design
3
ACCEPTED MANUSCRIPT
This retrospective, multicenter, cohort study involved four hospitals: National Cerebral and Cardiovascular Center (Osaka, Japan), Keio University Hospital (Tokyo, Japan), Kobe University Hospital (Hyogo, Japan), and Nagoya University Hospital (Aichi, Japan). This study was approved by the ethics committee of each institution and met
CR IP T
the guidelines of the Helsinki Declaration. Written informed consent was waived by the board because of the anonymous nature of the data. We enrolled patients aged ≥ 20 years undergoing thoracic vascular surgery with CPB between January 2010 and December 2012. Thoracic vascular surgery was divided into four groups: root surgery,
AN US
arch surgery, thoracic surgery, and combined surgery. Root surgery included mainly ascending artery replacement with or without additional surgeries such as valve replacement or coronary artery bypass grafting. Arch surgery included mainly total
M
arch replacement with or without additional surgeries such as valve replacement or coronary artery bypass grafting. Thoracic surgery included mainly descending thoracic
ED
aneurysm and thoracoabdominal aneurysm repairs. Combined surgery included mainly
PT
patients undergoing total arch replacement and thoracoabdominal aneurysm repair.
CE
Data collection
Data were collected from the hospitals’ electronic medical databases, which are
AC
uploaded to the Japan Adult Cardiovascular Database. This database consists of clinical information for all patients undergoing cardiovascular surgery in almost all hospitals in Japan, the details of which were described elsewhere15. The variables contained in the Japan Adult Cardiovascular Database are almost identical to those in the Society for Thoracic Surgeons' National Database, which collects detailed clinical information for patients undergoing cardiac surgery. However, the Japan Adult Cardiovascular
4
ACCEPTED MANUSCRIPT
Database lacks data on the administration of fibrinogen concentrate or cryoprecipitate; therefore, we retrieved this data from electronic medical records in each hospital and incorporated the data into the analyses.
CR IP T
Outcomes
The primary outcome was thromboembolic events (postoperative stroke, myocardial infarction, pulmonary embolism, and limb ischemia) during intensive care unit (ICU) stay. We defined thromboembolic events as: postoperative stroke, newly-emerged
AN US
paralysis of the central nerve system persisting more than 72 hours before discharge; postoperative myocardial infarction, newly-elevated cardiac biomarkers (creatine phosphokinase isoenzymes or creatine kinase MB isoenzyme greater than twice the
M
upper limit of the reference range, or elevated troponin); and postoperative pulmonary embolism diagnosed by lung perfusion scintigraphy or angiography; and limb ischemia
ED
and any complication caused by limb ischemia. The secondary outcome was 30-day
CE
PT
mortality.
Management of anticoagulation, and cardiopulmonary bypass
AC
This was an observational study, and anticoagulation and cardiopulmonary bypass management were neither controlled nor standardized. Usually, we administered 300– 400 U/kg of unfractionated heparin as a priming dose for anticoagulation for CPB to achieve an activated clotting time > 400 seconds. After CPB discontinuation, we used protamine to neutralize heparin. Postoperatively, patients were transferred to the ICU for monitoring and mechanical
5
ACCEPTED MANUSCRIPT
ventilation; patients were extubated when standard criteria were met. Usually, dynamic mechanical thromboprophylaxis (intermittent compression devices or pneumatic compression devices) or graduated compression stockings were used
CR IP T
postoperatively if there were no contraindications.
Transfusion practices
This was an observational study, and transfusion practices were neither controlled nor standardized. We usually administer red blood cells with a target hemoglobin level of 9–
AN US
10 g/dL, and this policy did not change through the study period. The administration of fresh-frozen plasma (FFP) and platelet concentrate were at the discretion of attending anesthesiologists
and
cardiac surgeons.
Usually,
FFP
was
administered
for
M
international normalized ratios > 1.5, and platelet concentrate was administered for platelet counts < 100 000 platelets/µL. In the National Cerebral and Cardiovascular
ED
Center, cryoprecipitate was transfused if fibrinogen was < 150 mg/dL. Usually, one pool of cryoprecipitate was prepared from 1440 ml of FFP, and the pool contained
PT
approximately 2 g of fibrinogen. In Nagoya University Hospital and Kobe University
CE
Hospital, fibrinogen concentrate (Fibrinogen HT; Japan Blood Products Organization, Tokyo, Japan) was administered if fibrinogen was < 150 mg/dL at weaning from CPB or
AC
when the attending surgeon believed that fibrinogen concentrate was required because of surgical bleeding. In Keio University Hospital, neither fibrinogen concentrate nor cryoprecipitate was used intra- or postoperatively; only FFP was administered, targeting a fibrinogen concentration of 120–150 mg/dl based on surgical hemostasis. As an antifibrinolytic agent, tranexamic acid was used at the anesthesiologist’s discretion.
6
ACCEPTED MANUSCRIPT
Statistics First, we eliminated patients with missing characteristics data (Fig 1). Then, we performed propensity score (PS) matching to balance patients’ backgrounds between the fibrinogen administration group (patients receiving fibrinogen concentrate or
CR IP T
cryoprecipitate) and no-fibrinogen group (patients not receiving fibrinogen concentrate or cryoprecipitate). PS was calculated to predict the probability of receiving fibrinogen concentrate or cryoprecipitate based on the predictors in Table 1 using a logistic regression model. We performed one-to-one patient matching between fibrinogen and
AN US
no-fibrinogen groups with the closest estimated PS within one caliper (≤ 0.2 of the pooled standard deviation of estimated logits) using the nearest-neighbor method. We estimated the balance in baseline variables using standardized differences, where >
M
10% was considered imbalanced16. We compared 30-day mortality and perioperative thromboembolic events between the groups using the 2 test.
ED
Second, we performed multivariate logistic regression analysis to confirm the results. To calculate odds ratios (OR) and 95% confidence intervals (95% CI) for 30-day mortality
PT
and thromboembolic events during ICU stay, variables with a P-value < .2 with
CE
univariate logistic regression analysis were selected for the multivariate logistic regression model with a forced entry of use of fibrinogen concentrate or cryoprecipitate.
AC
Regarding “age”, we calculated odds per 10 years and, as for “duration of operation” and “duration of CPB”, we calculated odds per 1-hour increase, for easier understanding, clinically. Multiple imputation was used for the multivariate logistic regression analyses to account for variables with missing data (n=1047). We performed all statistical analyses using IBM SPSS, version 22 (IBM Corp., Armonk, NY).
7
ACCEPTED MANUSCRIPT
Results
CR IP T
We enrolled 1047 patients during the study period. Postoperative outcome parameters in the four hospitals are shown in the supplementary table (Supplementary Table). To perform PS matching, we excluded 87 patients because of missing characteristics data, and analyzed data for 960 study patients. Our PS matching created 191 pairs of
AN US
patients with and without fibrinogen concentrate or cryoprecipitate administration (Fig. 1). The median amount of administered fibrinogen was 3 g (interquartile range: 2–4 g) in the fibrinogen group, which was calculated by assuming that one pool of
M
cryoprecipitate contained 2 g of fibrinogen.
Table 1 shows the baseline characteristics of the unmatched and PS-matched groups.
ED
After PS matching, standardized differences for the measured covariates, except degree of urgency, type of surgery, and hospitals, were < 0.1, suggesting that groups were
PT
generally well-balanced. Table 2 shows the proportion of 30-day mortality and the
CE
incidence of thromboembolic events between the PS-matched groups. We found no significant difference in the incidence of both outcomes between the groups. Even with
AC
1:2 and 1:3 matching, we found no significant difference in the incidence of both outcomes between the groups (data not shown). Next, to confirm these results, we performed multivariate logistic regression analysis. We identified several variables that differed between patients who experienced perioperative thromboembolic events (N=105) and those who did not (N=942) (Table 3). Multivariate analysis revealed that carotid vessel disease (OR, 2.25; 95% CI, 1.24–4.07;
8
ACCEPTED MANUSCRIPT
P=0.008) and duration of operation (per 1-hour increase: OR, 1.17; 95% CI, 1.05–1.29; P=0.003) were significantly associated with thromboembolic events. However, administering fibrinogen concentrate or cryoprecipitate was not associated with thromboembolic events. Similarly, age (per 10-year increase: OR, 1.43; 95% CI, 1.03–
CR IP T
1.97; P=0.031), preoperative resuscitation (OR, 14.07; 95% CI, 3.77–52.5; P<0.001), and duration of operation (per 1-hour increase: OR, 1.20; 95% CI, 1.02–1.40; P=0.028) were independent risk factors for 30-day mortality (Thirty-five of 1047 patients died). However, administering fibrinogen concentrate or cryoprecipitate was not associated
M
AN US
with 30-day mortality (OR, 0.44; 95% CI, 0.18–1.12; P=0.081) (Table 4).
ED
Discussion
This study investigated whether administering fibrinogen concentrate or to
treat
PT
cryoprecipitate
bleeding
is
associated
with
increased
postoperative
CE
thromboembolic events and/or improved mortality in patients undergoing thoracic vascular surgery with CPB. In this multicenter retrospective study using national
AC
registry data, PS matching revealed that administering fibrinogen concentrate or cryoprecipitate was associated with neither thromboembolic events nor 30-day mortality, which was confirmed by multivariate logistic regression models with a multiple imputation method for missing data. Previous randomized controlled trials have attempted to confirm the efficacy, but not the safety, of fibrinogen concentrate. 9 10 11, 12 These studies, given a low event rate of
9
ACCEPTED MANUSCRIPT
adverse events, are rarely powered to address safety issues. For example, death was reported in only 2/60 patients in the fibrinogen group, and none occurred in the control group in a recent randomized controlled trial12. Another randomized controlled trial reported thromboembolic events in 6/78 patients (7.7%) in the fibrinogen group and
CR IP T
10/74 patients (13.5%) in the placebo group11. A recent retrospective single-center cohort study suggested that perioperative administration of fibrinogen concentrate to bleeding patients undergoing cardiac surgery is not associated with increased incidence of major adverse cardiac and thromboembolic events and mortality17, which is consistent with
AN US
our findings. Contrasting with the previous study, which investigated patients undergoing cardiac surgery (with the exception of patients undergoing surgery of the ascending aorta17), we investigated patients undergoing thoracic aortic surgery, which
M
can be a more complex surgery with a higher risk of massive bleeding. We also enrolled patients from four different institutions, which potentially contributed to external
ED
validity18.
The ideal clinical situation in which to use fibrinogen concentrate or
PT
cryoprecipitate remains unknown. Theoretically, fibrinogen is the first coagulation
CE
factor to reach critically low levels during major blood loss and volume replacement, functioning both as a source of fibrin and as a mediator of platelet aggregation 4.
AC
Therefore, it is reasonable to hypothesize that rapid increases in fibrinogen levels would be more beneficial in patients undergoing complex surgery with a high risk of massive bleeding. To test this hypothesis, we investigated patients who underwent thoracic aortic surgery. Fibrinogen concentrate and cryoprecipitate were not used at all in one hospital, which may have worked in favor of PS matching because neither fibrinogen concentrate nor cryoprecipitate was used, even in severe cases. Our results confirmed
10
ACCEPTED MANUSCRIPT
the safety of fibrinogen concentrate or cryoprecipitate in our chosen patient population. Multivariate logistic regression analysis revealed that the OR of fibrinogen or cryoprecipitate use for 30-day mortality missed statistical significance (OR, 0.44; 95% CI, 0.18–1.12; P=0.081). In our additional logistic regression analysis dividing
CR IP T
fibrinogen concentrate and cryoprecipitate as explanatory variables, we found that fibrinogen concentrate may effectively reduce mortality (OR, 0.20; 95% CI, 0.05–0.84). Further studies with a higher study power may clarify the effect of using fibrinogen concentrate or cryoprecipitate.
AN US
We acknowledge several limitations of this study. First, because of the observational nature of this study, inference of causality is limited. Also, patients may have left the study, and misdiagnosis of rare adverse events is possible. For example,
M
spinal cord ischemia might present as stroke and be misidentified as a thrombotic complication. It is also possible that some events could have been missed because of the
ED
absence of standardized protocols. Second, although this study was based on multicenter data, the data were collected retrospectively. Despite using PS matching to
PT
reduce bias related to confounding, we cannot eliminate the possibility that there may
CE
be unobserved confounders such as differences in perioperative management between hospitals, practices, other blood product management systems, and the high likelihood
AC
that bleeding patients receive fibrinogen. Our PS matching results are generalizable only among patients in the PS range included in the paired analysis, and these results may not be applicable to patients with scores outside this range. However, we conducted multivariate logistic regression analysis using a multiple imputation method for missing data, and confirmed the robustness of the results. Third, our database lacked data on antifibrinolytic agents such as tranexamic acid used intraoperatively and
11
ACCEPTED MANUSCRIPT
anticoagulation therapy with antiplatelet or anticoagulation agents preoperatively, which may also have affected our results. Fourth, we lack laboratory data for perioperative fibrinogen levels or viscoelastic tests, which may be useful in guiding appropriate use of fibrinogen products, and this may have affected our results. Fifth, we
CR IP T
experienced 105 thrombotic events and 35 deaths. Considering this small number of deaths, the logistic regression analysis may have been affected by overfitting.
In conclusion, this study revealed that treatment with fibrinogen concentrate or cryoprecipitate was associated with neither thromboembolic events nor 30-day
AN US
mortality in patients undergoing thoracic aortic surgery.
Acknowledgments
M
This research was supported in part by the Research Grant for Regulatory Science of
ED
Pharmaceuticals and Medical Devices from Japan Agency for Medical Research and Development
PT
(AMED) and a grant-in-aid from the Takeda Science Foundation. We thank Jane Charbonneau,
CE
DVM, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
AC
Source of support:
This research was supported in part by the Research Grant on Regulatory Science of
Pharmaceuticals and Medical Devices from Japan Agency for Medical Research and development,
AMED, and a grant-in-aid from the Takeda Science Foundation.
12
ACCEPTED MANUSCRIPT
Conflicts of interest:
Yuichi Ueda received speaker honoraria and consulting fee from CSL Behring K.K. (Tokyo, Japan).
Shigeki Miyata received research support and speaker honoraria from Daiichi Sankyo Co., Ltd.
CR IP T
(Tokyo,Japan), Mitsubishi Tanabe Pharma Corporation (Osaka, Japan), and CSL Behring K.K.
(Tokyo, Japan). None of the other authors declare any conflicts of interest.
1.
AN US
References
Sniecinski RM, Levy JH: Bleeding and management of coagulopathy. J
2.
M
Thorac Cardiovasc Surg 142:662-667,2011
Murphy GJ, Reeves BC, Rogers CA, et al.: Increased mortality,
ED
postoperative morbidity, and cost after red blood cell transfusion in patients having cardiac surgery. Circulation 116:2544-2552,2007 Koch CG, Li L, Duncan AI, et al.: Morbidity and mortality risk associated
PT
3.
CE
with red blood cell and blood-component transfusion in isolated coronary artery bypass grafting. Crit Care Med 34:1608-1616,2006 Hiippala ST, Myllyla GJ, Vahtera EM: Hemostatic factors and replacement
AC
4.
of major blood loss with plasma-poor red cell concentrates. Anesth Analg 81:360-365,1995 5.
Karkouti K, Callum J, Crowther MA, et al.: The relationship between
fibrinogen levels after cardiopulmonary bypass and large volume red cell transfusion
in
cardiac
surgery:
an 13
observational
study.
Anesth
Analg
ACCEPTED MANUSCRIPT
117:14-22,2013 6.
Collins PW, Solomon C, Sutor K, et al.: Theoretical modelling of fibrinogen
supplementation
with
therapeutic
plasma,
cryoprecipitate,
or
fibrinogen
concentrate. Br J Anaesth 113:585-595,2014 Bolliger D, Szlam F, Molinaro RJ, et al.: Finding the optimal concentration
CR IP T
7.
range for fibrinogen replacement after severe haemodilution: an in vitro model. Br J Anaesth 102:793-799,2009 8.
Solomon C, Pichlmaier U, Schoechl H, et al.: Recovery of fibrinogen after
AN US
administration of fibrinogen concentrate to patients with severe bleeding after cardiopulmonary bypass surgery. Br J Anaesth 104:555-562,2010 9.
Rahe-Meyer N, Solomon C, Hanke A, et al.: Effects of fibrinogen
M
concentrate as first-line therapy during major aortic replacement surgery: a randomized, placebo-controlled trial. Anesthesiology 118:40-50,2013 Ranucci
M,
Baryshnikova
ED
10.
E,
Crapelli
GB,
et
al.:
Randomized,
double-blinded, placebo-controlled trial of fibrinogen concentrate supplementation
Rahe-Meyer N, Levy JH, Mazer CD, et al.: Randomized evaluation of
CE
11.
PT
after complex cardiac surgery. J Am Heart Assoc 4:e002066,2015
fibrinogen vs placebo in complex cardiovascular surgery (REPLACE): a double-blind
AC
phase III study of haemostatic therapy. Br J Anaesth 117:41-51,2016 12.
Bilecen S, de Groot JA, Kalkman CJ, et al.: Effect of Fibrinogen
Concentrate on Intraoperative Blood Loss Among Patients With Intraoperative Bleeding During High-Risk Cardiac Surgery: A Randomized Clinical Trial. JAMA 317:738-747,2017 13.
Wilhelmsen L, Svardsudd K, Korsan-Bengtsen K, et al.: Fibrinogen as a 14
ACCEPTED MANUSCRIPT
risk factor for stroke and myocardial infarction. N Engl J Med 311:501-505,1984 14.
van Hylckama Vlieg A, Rosendaal FR: High levels of fibrinogen are
associated with the risk of deep venous thrombosis mainly in the elderly. J Thromb Haemost 1:2677-2678,2003 Okita Y, Miyata H, Motomura N, et al.: A study of brain protection during
CR IP T
15.
total arch replacement comparing antegrade cerebral perfusion versus hypothermic circulatory arrest, with or without retrograde cerebral perfusion: analysis based on the Japan Adult Cardiovascular Surgery Database. J Thorac Cardiovasc Surg
16.
AN US
149:S65-73,2015
Austin PC: The use of propensity score methods with survival or
time-to-event outcomes: reporting measures of effect similar to those used in
17.
M
randomized experiments. Stat Med 33:1242-1258,2014
Fassl J, Lurati Buse G, Filipovic M, et al.: Perioperative administration of
ED
fibrinogen does not increase adverse cardiac and thromboembolic events after cardiac surgery. Br J Anaesth 114:225-234,2015 Bellomo R, Warrillow SJ, Reade MC: Why we should be wary of
PT
18.
AC
CE
single-center trials. Crit Care Med 37:3114-3119,2009
15
AC
CE
PT
ED
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
Fig. 1. Flow diagram of patient enrollment
Table 1. Baseline Patient Characteristics in the Unmatched and Propensity-Matched 16
ACCEPTED MANUSCRIPT
Groups Unmatched Groups Fib/Cryo
No
Standardi
Fib/Cryo
No
Standardi
administr
Fib/Cryo
zed
administr
Fib/Cryo
zed
ation
administr
Differenc
ation
administr
Differenc
(n=247)
ation
es, %
(n=191)
ation
es, %
(n=713)
(n=191)
CR IP T
variable
Matched Groups
Age (year)
67 (58-75)
69 (59-77)
-1.8
67 (59-76)
69 (60-77)
-2.7
Gender (%
163 (66.0)
487 (68.3)
-4.9
128 (67.0)
126 (66.0)
2.2
Height
163
164
-7.9
164
162
6.7
(cm)
(156-169)
(157-171)
Weight (kg)
60.3
61.6
(51-69)
(53-70)
124 (50.2)
345 (48.4)
64 (25.9)
185 (25.9)
177 (71.7)
535 (75.0)
27 (10.9)
males)
91 (47.6)
86 (45.0)
4.8
-0.1
46 (24.1)
51 (26.7)
-6.0
-7.6
136 (71.2)
129 (67.5)
8.0
77 (10.8)
0.4
19 (9.9)
22 (11.5)
-5.1
27 (10.9)
62 (8.7)
7.5
21 (11.0)
23 (12.0)
-3.3
2 (0.8)
5 (0.7)
1.3
4 (2.1)
2 (1.0)
8.4
28 (11.3)
61 (8.6)
9.3
24 (12.6)
23 (12.0)
1.6
8 (3.2)
14 (2.0)
8.0
6 (3.1)
5 (2.6)
3.1
19 (7.7)
62 (8.7)
-3.7
12 (6.3)
17 (8.9)
9.8
failure (%) Liver
M
ED
CE
dysfunctio
PT
(%) Renal
2.9
3.6
on (%) Diabetes
58.9
(51.5-67.0)
emia (%) Hypertensi
60.4
(52.0-69.5)
(%) Hyperlipid
(156-170)
AN US
Smoking
-10.5
(157-170)
n (%)
AC
Cerebrovas cular
accident (%)
Preoperati ve resuscitati on (%) COPD
17
ACCEPTED MANUSCRIPT
moderate to
severe 30 (12.1)
76 (10.7)
4.7
21 (11.0)
16 (8.4)
8.9
13 (5.3)
26 (3.6)
7.8
7 (3.7)
5 (2.6)
6.0
11 (4.4)
8 (1.1)
20.3
6 (3.1)
4 (2.1)
6.6
66 (26.7)
109 (15.3)
28.3
46 (24.1)
40 (20.9)
7.5
CCS (II-IV)
12 (4.9)
21 (2.9)
9.9
8 (4.2)
8 (4.2)
0
NYHA
182 (73.4)
437 (61.3)
26.7
137 (71.7)
133 (69.6)
4.6
good
182 (73.7)
M
(%) Peripheral
518 (72.7)
2.3
141 (73.8)
143 (74.9)
-2.4
moderate
60 (24.3)
185 (25.9)
-3.8
46 (24.1)
45 (23.6)
1.2
poor
5 (2.0)
10 (1.4)
4.8
4 (2.1)
3 (1.6)
3.9
Duration of
497
412
54.6
473
434
4.0
operation
(390-630)
(349-502)
(381-573)
(372-547)
Duration of
237
203
225
218
CPB (min)
(189-307)
(161-245)
(183-287)
(182-263)
Preoperati
9 (3.6)
25 (3.5)
0.7
7 (3.7)
7 (3.7)
0
elective
158 (64.0)
546 (76.6)
-27.9
127 (66.5)
112 (58.6)
16.3
urgent
20 (8.1)
41 (5.8)
9.3
10 (5.2)
20 (10.5)
-19.6
emergent
66 (26.7)
123 (17.3)
23.0
52 (27.2)
57 (29.8)
-5.7
salvage
3 (1.2)
3 (0.4)
8.8
2 (1.0)
2 (1.0)
0
vascular Old myocardial infarction (%) Congestive failure (%) Reoperatio n (%)
(II-IV)
AC
PT
CE
(min)
ED
LV function
ve
AN US
heart
CR IP T
disease (%)
54.9
9.4
shock
(%)
Urgency
18
ACCEPTED MANUSCRIPT
Arrhythmi
24 (9.7)
65 (9.1)
2.1
18 (9.4)
17 (8.9)
1.8
Lowest
24.5(22.1-
25.0
-18.2
24.5
24.0
-2.9
core
26.4)
(22.0-28.0)
(22.0-26.6)
(21.2-26.7)
a (%)
temperatur Type of cardiac surgery
CR IP T
e (˚C) Root
50 (20.2)
161 (22.6)
-5.7
38 (19.9)
34 (17.8)
5.4
arch
134 (54.3)
410 (57.5)
-6.6
102 (53.4)
118 (61.8)
-17.0
thoracic
51 (20.6)
120 (16.8)
9.8
43 (22.5)
29 (15.2)
18.8
combined
12 (4.9)
22 (3.1)
9.1
8 (4.2)
10 (5.2)
-4.9
NCVC
103 (41.7)
336 (47.1)
-10.9
99 (51.8)
65 (34.0)
36.6
Keio
0 (0)
198 (27.8)
-87.7
0 (0)
37 (19.4)
-69.3
30 (12.1)
135 (18.9)
-18.8
25 (13.1)
60 (31.4)
-45.2
114 (46.2)
44 (6.2)
67 (35.1)
29 (15.2)
-47.1
University Kobe University Nagoya
102.1
ED
Hospital
M
Hospital
AN US
Hospital
University Hospital
PT
Numerical data are presented as medians (1st quartile–3rd quartile). Categorical data are presented as number (%). Abbreviations: Fib, fibrinogen concentrate; Cryo, cryoprecipitate; COPD, chronic
CE
obstructive pulmonary disease; CCS, Canadian Cardiovascular Society functional classification; NYHA, New York Heart Association; NCVC, National Cerebral
AC
Cardiovascular Center
Table 2. Postoperative outcome parameters in the propensity-score-matched groups 19
ACCEPTED MANUSCRIPT
characteristic Thrombosis
Fibrinogen/cryoprecipitate
No Fibrinogen/cryoprecipitate
P
group (n=191)
group(n=191)
27 (14.1)
28 (14.7)
0.884
6 (3.1)
10 (5.2)
0.307
(%) 30-day mortality (%)
CR IP T
Categorical data are presented as number (%).
Table 3. Multivariate logistic regression analysis of patients' thrombosis events in the intensive care units Univariate analysis
Multivariate analysis
OR
95% CI
Fibrinogen/
1.65
1.08-2.52
0.021
95% CI
P
1.22
0.76-1.95
0.408
1.10
0.95-1.27
0.192
1.12
0.95-1.32
0.188
Gender (male)
1.55
0.97-2.46
0.064
1.22
0.71-2.07
0.474
Smoking (missing 1)
1.59
1.06-2.40
0.026
1.26
0.78-2.03
0.342
DM
1.66
0.95-2.91
0.075
1.37
0.74-2.53
0.323
Carotid vessel disease
2.29
1.31-4.00
0.004
2.25
1.24-4.07
0.008
1.76
0.94-3.30
0.078
1.37
0.70-2.70
0.362
1.51
0.88-2.59
0.139
1.20
0.66-2.18
0.560
1.75
0.76-4.02
0.190
1.03
0.40-2.68
0.955
1.39
0.86-2.24
0.178
1.02
0.59-1.78
0.943
1.14
1.07-1.22
<0.001
1.17
1.05-1.29
0.003
1.19
1.06-1.35
0.004
0.99
0.78-1.14
0.559
cryoprecipitate use Age (per 10 years
PT
Peripheral vascular
ED
severe
M
increase)
COPD moderate to
P
OR
AN US
covariates
disease (missing 51)
CE
Old myocardial
infarction (missing 2)
AC
Reoperation
Duration of operation (per 1 hour increase) Duration of CPB (per 1 hour increase) (missing 77)
Type of cardiac surgery (missing 31) Root
1
1
(reference)
(reference) 20
ACCEPTED MANUSCRIPT
Arch
1.00
0.60-1.67
0.999
0.76
0.44-1.30
0.307
Thoracic
0.86
0.44-1.68
0.660
0.57
0.27-1.21
0.142
Combined
2.80
1.14-6.90
0.025
1.75
0.68-4.54
0.249
Abbreviations: OR, odds ratio; CI, confidence interval; DM, diabetes mellitus; COPD, chronic obstructive pulmonary disease; CPB, cardiopulmonary bypass; missing (n),
AC
CE
PT
ED
M
AN US
CR IP T
missing data for the number of patients (n)
21
ACCEPTED MANUSCRIPT
Table 4. Multivariate logistic regression analysis of 30-day mortality Multivariate analysis
OR
95% CI
P
OR
95% CI
P
Fibrinogen/
1.11
0.52-2.33
0.794
0.44
0.18-1.12
0.081
1.22
0.94-1.58
0.133
1.42
1.03-1.95
0.03
Renal failure
2.09
0.84-0.18
0.109
2.02
0.71-3.35
0.16
Carotid vessel
2.59
1.10-6.11
0.029
2.08
0.74-5.45
0.17
16.6
6.3-43.9
<0.001
13.0
3.59-47.3
<0.001
2.51
1.01-6.24
0.047
2.54
0.90-7.16
0.08
1.23
<0.001
1.20
1.02-1.43
0.02
<0.001
1.23
0.91-1.65
0.17
CR IP T
covariates
M
Univariate analysis
cryoprecipitate use Age (per 10 years
AN US
increase)
disease Resuscitation (missing 1) COPD moderate to severe Duration of
ED
operation (per 1 hour increase) (per 1 hour
1.38
1.17-1.63
PT
Duration of CPB
1.13-1.34
77)
CE
increase) (missing Urgency
AC
elective
1
1
(reference)
(reference)
Urgent
1.05
0.24-4.59
0.946
0.67
0.11-4.12
0.66
Emergent
2.16
1.02-4.59
0.045
1.25
0.77-2.05
0.65
Salvage
18.7
3.23-108.0
<0.001
1.59
0.17-14.7
0.68
0.16-0.96
0.04
Type of cardiac surgery (missing 31) Root Arch
1
1
(reference)
(reference)
0.51
0.22-1.14 22
0.102
0.39
ACCEPTED MANUSCRIPT
Thoracic
0.84
0.36-2.31
0.839
0.81
0.26-2.56
0.72
Combined
1.26
0.27-5.93
0.774
0.77
0.33-1.81
0.76
Abbreviations: CI, confidence interval; P, p-value; COPD, chronic obstructive pulmonary disease; CPB, cardiopulmonary bypass; missing (n), missing data for the number of
AC
CE
PT
ED
M
AN US
CR IP T
patients (n)
23