正常胸部 亀田 徹，他 気胸以外に Lung sliding signが seashore 消失する疾患 Fig. 1. Longitudinal ultrasound images from the linear probe of the uninjured left anterior chest in a 70-year-old man who presented with right lateral chest pain after falling from a ladder. ! a: B-mode image show 気胸s the pleural line (white arrows) and a comet-tail artifact (arrowheads). Lung sl外idi傷ng性 w気as 胸obのse超rve音d 波診断 with real-time images. b: M-mode image shows seashore sign. ! c: Color Doppler image shows presence of the Doppler signal on artifacts. R: rib, IM: intercostal muscle 呼吸停止や減弱, 胸膜の癒着, 合は，胸壁には呼吸性の動揺がないので平行線が多 Acute respiratory distress syndrome（以下ARDS）23,37)，無肺挫傷, COPD, 肺線維症, 数描出されるが，pleural line 及びその深部に描出さ 気肺38)，片肺挿管5,38) では気胸がなくてもlung sliding れる多重反射よるアーチファクトには呼吸性の動揺 がみられないことが挙げられる。 ARDS, 無気肺, 片肺挿管 が起こるので砂浜様に描出される。Lichtenstein ら 2）Comet-tail artifact（B line） は胸壁にあたる平行線の部位を砂浜に打ち寄せる波 Comet-tail artifact は，ある物質と，それを取り囲 に擬え，seashore sign として報告している 30)（Fig. む音響インピーダンスの極端に異なる物質の間で起 lung sliding （-） barcode 1b）。一方，気胸の場合は pleural line より深部のアー こる多重反射で，はっきりとした輪郭を持った線状 チファクトにも呼吸性の動揺がみられないので，平 高輝度エコーのことを指す39)。肺では，胸膜下の肺 行線が描出され seashore sign はみられない 30)（Fig. 胞内空気と，肥厚した小葉間中隔や増加した血管外 Fig. 2. Longitudinal ultrasound images from the linear probe of the right anterior chest with pneumothorax in the same patient. 2b）。まa: B た -m カ od ラ e i ー ma パ ge ワ sho ー w ドs th プ e p ラ leu ， ra カl li ラne ー (w ドhite プ ar ラ row を s) 利 and no com 水 et 分-ta とil a の rtif 間 ac でt. L 起 un こ g s る lid と ing さ w れ as る n3o0)t o （ bFsiegr.v e 1 d a w ） it 。h このアー real-time images. 用すれ b: ば M ， -m 気 ode 胸 im が ag な e d い oe 場 s n 合ot はshpolw e u sreaals hlionre e sig よ n. り深部の チファクトは pleural line から画面の深部端まで伸 c: Color Doppler image shows absence of the Doppler signal on the pleural line and the underlying artifacts. アーチファクトには呼吸性の動揺が起こるのでカ び，lung slidingがある場合にはそれにあわせて移動 R: rib, IM: intercostal muscle ラーはのるが（Fig. 1c），気胸がある場合にはカラー することが特徴である30)。小葉間中隔の肥厚に由来 はのらない 5,14,33)（Fig. 2c）。ただし，これを診断目 するとされるcomet-tail artifactは，X線所見のKerley 観 的 察 で 部 利 位 用 に す 気 る 胸 場 が 合 あ ， れ ゲ ば イ ， ン 肺 を 表 適 面 切 ま に で 調 超 整 音 し 波 ないが と届 判 で の， B 30) p l lienueral に lin 相 eの 当 異 する常 とが さ観 れ 察 ， さ B lれ in れ e ば と 気 も 胸 呼 が ば な れるいこ 。 かな 断 い を の 誤 で る ， 可 肺 能 表 性 面 が で あ 起 る こ の る で c ，om 著 et 者-ta らil a の rti 見 fac 解 tは で み は ら ，B とが C 言 om え et-る taiは l aず rti で fa あ ct る は 。 正 た 常 だ 肺 こ で れ も ら 観 の 察 疾 さ 患 れ で る は が co ， m- あ れな モ い ー （ ド 25) F で ilgu. n2ga ） sl 。 idi こ ng のア を ー 確 チ 認 フ の ァ 上 ク ，静ト 止が 画あ とれ しば て， 保観 存 et- まtai りl ar 目 tif 立act た も な 増 い 加す 。 る 水 の 分 で 貯 ， 留 co で m 小et- 葉tail 間 ar 中 tif 隔act の を 肥 補 厚 完 す 察部 す 位 る に 目 気 的 胸 に が と な ど い め と た 断 方 定 が す 無 る 難 こ と と 考 が え で る き 。 る。しか す るるサ 心 イ 原 ン 性 で 肺 は 水 な 腫 い 39) か ， も 結 し 合 れ 組 な 織い でが， 肥 観 厚 察 す が る 容 肺 易 線 で 維 し健常 注 肺 意 や す 39,40) 39,41) C べ O き PD 点 で と は し こ て の ， ア 肺 ー の チ 呼 フ 吸 ァ 性 ク 変 ト 動 が の 確 消 認 失やで 減 あ 症り，今 ， 後 ac そ ute の lu有 n 用 g i 性 njuに ry つ （ い 以ての 下 A検 LI討 ）が期 や A待 R さ DSれる。 ， きな 少 い ， こ 胸 と 膜 が の癒あ 着る 等の に で よ ， り こ ， の 肺 ア 挫 ー 傷1チ 4,34 フ ) ァ ，chク ro ト nic が obみ str ら uc- 4） 肺 L 挫un 傷 32) g puls で eはこのアーチファクトは重症度に合わ れ ti な ve い pulだ m け on で ary は di 気 sea胸 se と （ 断 以 定 下 す C る OP こ D と ）35,36)はで ， き 肺 な 線 い 維 43) 症 23) 。， せ片 て肺 増換 加気 しや ，無 多呼 く 吸 な ， る 胸 と 膜 癒 の 合 癒 す 着 る40-42) 等で肺 。 に動きがな その他注意すべき点として，外傷性気胸に合併し く，pleural line に沿った水平方向の動きである lung 134 JJAAM. 2012; 23: 131-41 やすい皮下気腫でもcomet tail artifactがみられること sliding がみられない時，心拍動が肺に伝わっている である。皮下気腫ではcomet-tail artifactはpleural line 様子が，pleural line に対して垂直方向の動きとして より腹側から起こり，量が多ければpleural lineの描 観察され，Lichtenstein らにより lung pulse と命名さ 出を妨げるので，皮下気腫の存在部位をpleural line れた 38)。気胸で肺が胸壁に接していなければ，心拍 と誤認する可能性がある。肺表面由来の comet-tail 動は胸腔内の空気を介して壁側胸膜に伝わらないの artifact を皮下気腫によるものと区別するためには， で lung pulse はみられない。なんらかの理由で lung 超音波プローブを体幹の長軸にあて肋間を描出し， sliding が確認できないときに，lung pulse があれば 肋間深部に描出されるpleural lineを正確に同定する 肺が胸壁に接していることを表し，観察部位の気胸 必要がある30)。 を除外することができるとされる38)。 3）Pleural line abnormalities 5）Lung point 肺線維症ではpleural lineが分断され不整で厚くな ある程度の空気が胸腔に存在すると，仰臥位では る44)。ALI/ARDSでも同様の異常が観察される41)。病 空気は腹側に集まり，肺は背側へ移動し，壁側胸膜 的変化が肺で起こっても胸壁側では起こらないの と臓側胸膜が接する部位と接しない部位との境界が 日救急医会誌. 2012; 23: 131-41 135
Muller et al. Critical Care 2012, 16:R188 Page 4 of 7 http://ccforum.com/content/16/5/R188 Table 1 Characteristics of the general population and comparison between responders and non-responders at baseline (before fluid challenge) All patients Responders Non-responders P-value (n = 40) (n = 20) (n = 20) Age, years 63 (56, 70) 61 (49, 70) 66 (53, 75) 0.58 Weight, Kg 72 (65, 77 67 (63, 76) 76 (63, 88) 0.14 Height, cm 169 (164, 173) 170 (162, 176) 168 (160, 173) 0.38 APACHE II score 17 (14, 23) 18 (14, 29) 14 (11, 21) 0.30 Heart rate, bpm 101 (91, 116) 101 (91, 125) 103 (79, 121) 0.78 Mean arterial pressure, mmHg 71 (66, 77) 70 (61, 88) 72 (65, 87) 0.56 LVEF, % 55 (50, 60) 55 (50, 60) 55 (47, 60) 0.41 Velocity time index, cm 16 (14, 18) 14 (12, 16) 17 (15, 21) < 0.01 E velocity, cm/s 75 (70, 80) 65 (53, 76) 82 (75, 93) < 0.01 E/A velocity ratio 0,9 (0.8, 1.1) 0,8 (0,6, 1,1) 1,0 (0,8, 1,4) < 0.01 Ea velocity, cm/s 12 (10, 13) 12 (9, 14) 11 (9, 15) 0.79 E/Ea velocity ratio 6 (5, 8) 5 (5, 10) 7 (5, 8) 0.40 cIVC, % 34 (16, 64) 64 (28, 100) 19 (5, 35) < 0.01 Data are expressed in medians with 5th and 95th percentiles. APACHE: Acute Physiology and Chronic Health Evaluation; cIVC, collapsibility index of the inferior vena cava. The first explanation for these imperfect results is that, been shown that there is a good correlation between high as previously suggested, cIVC is a dynamic preload index. cIVC value and low CVP value [21,22,42]. A low CVP In contrast with findings reported in mechanically venti- value (< 7 mmHg) could be considered a good indicator of lated septic patients, dynamic parameters have been fluid responsiveness , corresponding to high values of shown to be ineffective to predict fluid responsiveness in cIVC (specificity = 80%). In contrast, lower values of cIVC spontaneous breathing patients [6,7]. Spontaneous ventila- values are poorly predictive, corresponding to higher tion implies a very wide range of breathing patterns. In values of CVP [8,9]. patients with spontaneous ventilation, respiratory varia- The conditions of measurement of cIVC could be dis- 下大静脈径（ tions are highly variable from one cycle to another in a cussed. In the present study, the IVC diameter was mea- given patient and between different patients. Then, influ- sured by M mode at 2 or 3 cm from the right atrium, as ence of breathing pattern on cIVC is also variable. The described in previous studies [17,18,22]. However, Wal- present results indirectly confirm that spontaneous breath- E- lace et al. [4 F 3] re A cently S show T ed tha） t in spontaneously ing is a natural limit for the use of a dynamic parameter. breathing healthy volunteers, variations of IVC diameter Because previous studies have reported a good correla- were significantly lower when recorded closed to the tion between cIVC and blood volume removal during hemodialysis [17,24] or during blood donation , the inability of cIVC to predict fluid responsiveness may be • フラン su スのニ rprising in ー sp ム大学病院 ontaneously IC breath U in （ g 16 pati 床）に入室 ents with ACF. However, monitoring blood volume during blood removal した急性循環不全の患者（ is not the same as predicting n=4 fluid r 0） esponsiveness. It has >40%ならまず輸液 • 敗血症（ Table 2 60%） Causes , of 出血（ acute 28%） circulatory , 脱水（ failure 13%） Pathology Number of patients (%) Sepsis 24 (60) Intra-abdominal infection 10 • IVCの呼吸性変動と6% HES（ボルベン®） Pulmonary infection 9 500ml 15分で投与に対する輸液反応性を調査 Pyelonephritis 5
Bleeding 11 (28) Postoperative 7 Trauma 4 • 40%をCut offにすると, 感度 70%, 特異度 80% Dehydration 5 (13) Figure 1 Individual values of inferior vena cava collapsibility (cIVC) (%) after infusion of 500 mL of HES. The best cutoff value The total percentage is different from 100% because specific percentages is 40%. were rounded. Muller L, Bobbia X, Toumi M, et al. Respiratory variations of inferior vena cava diameter to predict fluid responsiveness in spontaneously breathing patients with acute circulatory failure: need for a cautious use. Crit Care. 2012 Oct 8;16(5):R188.
FFPとRCCの比率 RatioofBloodProductsAffectsMortalityinTrauma 100 18.5 • USの軍人病院で24時間以内にRCC 10単位以上を輸血 90 80 14 された患者（n=246）の後ろ向き研究. 70 60 • 外傷で凝固異常を認める患者では, FFPを十分投与するRatio of 50 Blood Products Affects Mortality in Trauma ことが生存退院への独立した因子であった. 40 11.5 70 Table 1. Severe 30 (AIS scores of 4 and 5) thoracic injuries were 65% more common in the low ratio group compared with in7 the 60 6 medium and 20 high groups. All vital signs and laboratory re- 50 sults were comparable, 4 2 10 except for hemoglobin, which 2.5 was 40 significantly lower1 in the 1 1 0.5low ratio group compared with in rtality 34% the medium0 and high groups. 30 Mo Low n=20 Medium n=18 High n=31 In the first 24 hours of admission, the rate per hour of 19% 20 crystalloid and † RBC units 92.5 a administered 78 a was less in 37 b Hemorrhage % the high 10 ratio group Sepsis % compared5 with in the medium 6 and low 19 groups MOF % 0 11 13 0 (Table 2). The total amount and rate per hour of plasma as (Low) 1:8 (Medium) 1:2.5 (High) 1:1.4 Airway/Breathing % 0 6 8 well as the rate per hour of FWB was higher in the medium CNS % 2.5 0 23 Plasma:RBC Ratio Groups and high Time to ratio death (hrs)2* groups ( p 2 (1 Ͻ 0.001). The low ratio group did – 4) a 4 (2-16) b 38 (4 – 155) c Fig. 2. not Comparison of receive the number aPLTs, and percentage which of the were primary only cause of used death for all in of 27% the deaths of in patients. Fig. 1. Percentage Borgm mortality an MA, S associated pine with lla P low, C, Perki medium, ns JG and , e hight al. The ratio of blood products transfused affects morta each lity plasma in to RBC ratio group. plasma to RBC ratios patients r transfusedece at iving mas admission.sive tra Ratios nsf areusion Numbers ona median t a c column omba representst supp absolute ort Cryoprecipitate h number os that p was ita died l. used from J T eachrau morem cause a in . listed. 20 the 0 When 7 twoO high ct;6 causes 3 ratio(4 were ):8 listed 05 group for - a( 1p3.Ͻ patient, they were counted as 0.5. 2Data presented in hours as median (interquartile range); *Mann-Whitney U test; †Chi Square test. Values with different superscripts ratios per group and include units of fresh whole blood counted (a, b,both c) are 0.01), significantly though different (p Ͻ given 0.05). at a higher rate in the medium and low as plasma and RBCs. ratio groups ( p Ͻ 0.001) and was only used in 51% of the Table 3 Comparison of Mortality Rates of Alternative Patient Cohorts and Plasma to RBC Ratios patient population (Table 2). Plasma to RBC Ratio (Range) Low Ratio (0:22–1:4) Medium Ratio (1:3.9–1:2.1) High Ratio (1:2–1:0.59) Primary analysis* Nonsurvivors in 65%a nthe ϭ 31 low and medium 34%b n ϭ ratio 53 groups died 19%c n ϭ 162 sion model initially used all variables associated with Excluding significantly thoracic trauma sooner 57%a than n ϭ 23 ϭ 26 those in the high 29%b n ϭ 36%b n ϭ ratio 48 50 group (Fig. 19%b n ϭ 152 Excluding neurotrauma 62%a n 15%c n mortality with a p value ϭ 145 Ͻ0.2. Variables were removed if Excluding Excluding 2). whole rFVIIa Median blood† time of death 66%a n ϭ 38 69%a n ϭ 26 measured in hours 27%b n ϭ 59 38%b n ϭ 39 from admission 19%b n ϭ 149 significant colinearity was measured by Pearson’s correlation 15%c n ϭ 100 Values with to the different hospital superscripts (a, b, c)was are 2 hours significantly (interquartile different ( p Ͻ 0.05) (Chi range, Square Test). 1– 4) in the coefficient or variance inflation. * See low Methods group Section. † Ratio calculated as FFP:RBC and units. 4 hours (interquartile range, 2–16) in the All continuous nonparametric data are described as me- medium group, compared with 38 hours (interquartile range, dian (interquartile range) or median (interquartile range) whole blood, 4 but –155) rather in called the for a high much ratio greater group percentage ( p Ͻ plastin 0.001). time (PTT) was 1.5 times normal, or after 10 RBC [mean] if the median value is zero. Mann-Whitney U test of and RBC units.35 Such protocols recommended that FFP only units were transfused. Additionally, these massive transfusion be transfused if The prothrombin relationship time (PT) or partialbetween thrombo- plasma protocols to called RBC for 1 ratios unit of FFP trans- to be given for every 4 to Kruskal Wallis tests were used for comparisons of continuous fused and overall mortality remained in the alternative anal- data. All categorical data were compared with 2 or Fisher’s Volume 63 • Number 4 809 yses performed. Differences in mortality remained significant exact test as appropriate. Statistical analysis was performed in the high, compared with in the low, ratio group when with SPSS 14.0 (Chicago, IL). Significant differences were patients with thoracic and head trauma were individually determined at p Ͻ 0.05. removed from the analysis (Table 3). The relationship between RESULTS plasma to RBC ratios transfused and overall mortality also Between November 2003 and September 2005, 5,293 remained when only stored FFP and RBCs (FWB units not patients were admitted to the CSH in Baghdad. The JTTR included) were used to calculate the ratio, as well as when identified 246 (4.6%) patients who received massive transfu- patients who were treated with rFVIIa were excluded (Table 3). sion. Penetrating injuries occurred in 232 of 246 (94%) of Table 4 indicates that many of the admission vital signs, these patients. Three patients were female. The median age of laboratory values, and Injury Severity Scores, in addition to the patients studied was 24 years (interquartile range, 21–30). the ratio of plasma to RBCs, were associated with overall The median ISS was 18 (interquartile range, 16 –25). The mortality. Table 5 reveals that the plasma to RBC ratio was combat support hospital length of stay for patients was a independently associated with overall survival (odds ratio median of 2 (1– 6) days. The median time from admission to 8.6, 95% confidence interval 2.1–35.2) and that both base the CSH to evacuation to Germany for US patients was 1 day deficit and AIS for head and neck were independently asso- (1–2 days). The overall mortality was 28%. Median plasma to ciated with decreased overall survival upon logistic RBC transfusion ratios for survivors versus nonsurvivors regression. were 1:1.6 (1:1.3–1:2.2) and 1:2.3 (1:1.4 –1:5.1), respectively Figure 2 displays the primary causes of death in each ( p Ͻ 0.001). Median plasma to RBC ratios were 1:8, 1:2.5, ratio group. The percentage of deaths from hemorrhage and 1:1.4, and are defined as low, medium, and high ratios was lower in the high ratio group (11.5 of 31; 37%), respectively, ( p Ͻ 0.001). The low, medium, and high ratio compared with in the low ratio group (18.5 of 20; 92.5%) groups had plasma to RBC ratio ranges of 0:22–1:4, 1:3.9 – ( p Ͻ 0.001). This represents an absolute reduction of 55% 1:2.1, and 1:2–1:0.6, respectively. As the ratio of plasma to and a relative reduction of 60%. There were fewer hem- RBC increased, mortality significantly decreased (Fig. 1). orrhagic deaths in the high ratio group compared with in The mortality of low, medium, and high groups were 65%, the medium ratio group ( p Ͻ 0.05). Likely reflecting the 34%, and 19%, respectively ( p Ͻ 0.001). increased survival time, multiorgan failure deaths were Descriptive statistics for severity of injury, admission more frequent in the high ratio group compared with in the vital signs, and laboratory values for the three groups are in low ratio group. Volume 63 • Number 4 807
Wade et al. International Journal of Emergency Medicine 2011, 4:35 Page 3 of 6 http://www.intjem.com/content/4/1/35 80 70 60 ) 50 (% Mortality (%) lity 40 ta Filter r Mo 30 20 10 0 30 30.5 31 31.5 32 32.5 33 33.5 34 34.5 35 35.5 36 36.5 37 37.5 38 38.5 39 39.5 40 40.5 41 41.6 42.2 42.9 44 Temperature (C) Figure 1 Percent mortality at various body temperatures in patients with traumatic injuries as determined from the NTDB, n = 621,081. differences in mortality rates between the civilian and temperature may be a delineating physiological factor in military groups for hypothermia and normal tempera- assessing civilian patient status on presentation. ture groups. The increase in mortality in patients with In military patients, there were indices of increased hyperthermia was greater in the civilian patients than severity of injury in patients with hypothermia and the military. hyperthermia (Table 3). The injury severity score (ISS) For the civilian population, there were significant mor- was increased, and the trauma revised ISS (TRISS) was tality differences between the patients within the normal reduced in these groups. The RTS was not different body temperature range and those with hypothermia between groups, although with hypothermia, a signifi- and hyperthermia (Table 2). Those patients with cantly lower mean systolic blood pressure was noted. hypothermia and hyperthermia had significant increases Patients with hypothermia or hyperthermia spent a Wade et al. in indice体温管理は重要 International Journal of Emergency Medicine 2011, 4:35 Page 3 of 6 http://www.intjem.com/content/4/1/35 s of injury, more ventilator days, and greater greater number of days on ventilation and in the ICU. lengths of stay in the intensive care unit (ICU) and in There was no significant difference in the length of hos- the hospital. Of note, there were no differences in phy- pital stay. On the whole, military patients, while more 80 siological indices such as systolic pressure and revised severely injured and younger, showed similar mortality 70 trauma score (RTS), suggesting that differences in body differences as those seen for civilian patients in the 60 effects of hypothermia and hyperthermia. ) 50 In the military population, the Barell index for trau- (% Mortality (%) Table 1 Percent mortality in civilian and military patients matic brain injury (TBI) was recorded and the anatomi- lity 40 ta Filter r Study population Hypothermia Normal Hypothermia cal injury score (AIS) for brain trauma was coded. The Mo 30 (< 36°C) (36°-38°C) (> 38°C) incidence of moderate or severe brain injury with the 20 Civilian 12.3% (47/382)* 2.3% (84/3,619) 14.1% (13/92)* Barell index in patients was 55% in the hypothermia 10 Military 11.0% (29/263)* 1.7% (66/3,804) 3.7% (12/327)*+ group, 36% in the normal group, and 49% in the *Significantly different (p < 0.05) from normal. hyperthermia group. With an AIS > 2, there was a 53% 0 + 30 30.5 31 31.5 32 32.5 33 33.5 34 34.5 35 35.5 36 36.5 37 37.5 38 38.5 39 39.5 40 40.5 41 41.6 42.2 42.9 44 Significantly different (p < 0.05) from civilian. rate of brain injury for the hypothermia group, 35% for Temperature (C) Figure 1 Percent mortality at various body temperatures in patients with traumatic injuries as determined from the NTDB, n = Wade CE, Salinas J, Eastridge BJ, et al. Admission hypo- or hyperthermia and survival 621,081. after trauma in civilian and military environments. Int J Emerg Med. 2011 Jun 23;4(1):35. differences in mortality rates between the civilian and temperature may be a delineating physiological factor in military groups for hypothermia and normal tempera- assessing civilian patient status on presentation. ture groups. The increase in mortality in patients with In military patients, there were indices of increased hyperthermia was greater in the civilian patients than severity of injury in patients with hypothermia and the military. hyperthermia (Table 3). The injury severity score (ISS) For the civilian population, there were significant mor- was increased, and the trauma revised ISS (TRISS) was tality differences between the patients within the normal reduced in these groups. The RTS was not different body temperature range and those with hypothermia between groups, although with hypothermia, a signifi- and hyperthermia (Table 2). Those patients with cantly lower mean systolic blood pressure was noted. hypothermia and hyperthermia had significant increases Patients with hypothermia or hyperthermia spent a in indices of injury, more ventilator days, and greater greater number of days on ventilation and in the ICU. lengths of stay in the intensive care unit (ICU) and in There was no significant difference in the length of hos- the hospital. Of note, there were no differences in phy- pital stay. On the whole, military patients, while more siological indices such as systolic pressure and revised severely injured and younger, showed similar mortality trauma score (RTS), suggesting that differences in body differences as those seen for civilian patients in the effects of hypothermia and hyperthermia. In the military population, the Barell index for trau- Table 1 Percent mortality in civilian and military patients matic brain injury (TBI) was recorded and the anatomi- Study population Hypothermia Normal Hypothermia cal injury score (AIS) for brain trauma was coded. The (< 36°C) (36°-38°C) (> 38°C) incidence of moderate or severe brain injury with the Civilian 12.3% (47/382)* 2.3% (84/3,619) 14.1% (13/92)* Barell index in patients was 55% in the hypothermia Military 11.0% (29/263)* 1.7% (66/3,804) 3.7% (12/327)*+ group, 36% in the normal group, and 49% in the *Significantly different (p < 0.05) from normal. hyperthermia group. With an AIS > 2, there was a 53% +Significantly different (p < 0.05) from civilian. rate of brain injury for the hypothermia group, 35% for
トランサミン® • 40カ国, 274施設で行われたRCT • 対象: 受傷後8時間以内, 大量出血もしくはハイリスク患者 • 大量出血: sysBP <90mmHg or HR >110bpm or その両方 • 出血リスクが高い: 治療医の判断 • トランサミン（n=10960） vs. placebo（n=10115） • トランサミン 1g 10分かけて静注 → 8時間後に再投与 • Primary outcome: 受傷後4週間以内の死亡（出血死, 血管閉塞, 多臓器不全, 頭部外傷など死因も調査） CRASH-2 trial collaborators, et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet. 2010 Jul 3;376(9734):23-32.
死亡率減少 14.5% vs. 16.0% （RR 0.91, 95% CI 0.85-0.97, p=0.0035） 血管閉塞（心筋 塞, 脳 塞, PE）には差がなかった.
Articles 3時間以内に… Tranexamic acid allocated Placebo allocated Risk ratio (95% CI) Time to treatment (h) ≤1
181/3272 (5·5%) 227/3250 (7·0%) 0·79 (0·66–0·96) χ²=0·923; p=0·34 All deaths
489/10 060 (4·9%) 574/10 067 (5·7%) 0·85 (0·76–0·96) Two-sided p=0·0077 0·6 0·8 1 1·2 1·4 1·6 Tranexamic acid better Tranexamic acid worse Figure 1: Mortality due to bleeding by subgroups CRASH-2 trial collaborators, et al. The importance of early treatment with tranexamic acid in bleeding trauma patients: an exploratory analysis of the CRASH-2 randomised controlled trial. Lancet. 2011 Mar 26;377(9771):1096-101, 1101.e1-2. independence of any observed treatment interactions we bleeding was signifi cantly reduced with tranexamic acid. ran a logistic model including all possible interactions in 489 of 10 060 (4·9%) patients died because of bleeding in the four prespecifi ed baseline characteristics and the tranexamic acid group versus 574 of 10 067 (5·7%) in treatment subgroups. the placebo group (RR 0·85, 95% CI 0·76–0·96; A logistic regression was estimated with death due to p=0·0077). We noted no signifi cant eﬀ ect on the risk of bleeding as the dependent variable and treatment group death for all other (non-bleeding) causes combined and time to treatment as explanatory factors. We included (table table 1). an interaction parameter to allow for a proportional change Ta Table 2 shows the baseline characteristics of patients in the odds ratio (OR) as time to treatment increases. ORs according to time to treatment. Fi Figure 1 shows the and 95% CIs were estimated for diﬀ erent times to results of the subgroup analyses for death due to treatment. CIs were calculated with a logistic model with bleeding. Time to treatment was unknown in nine time as a continuous term and an interaction term between participants. Treatment given 1 h or less from injury time and tranexamic acid. We also ran a model with an signifi cantly reduced the risk of death due to bleeding interaction term for time to treatment squared to allow for (198/3747 [5·3%] in tranexamic acid group vs 286/3704 a non-constant proportional change in the OR. [7·7%] in placebo group; RR 0·68, 95% CI 0·57–0·82; The trial is registered as ISRCTN86750102, p<0·0001). Treatment given between 1 and 3 h also ClinicalTrials.gov NCT00375258, and South African reduced the risk of death due to bleeding (147/3037 Clinical Trial Register/Department of Health DOH-27- [4·8%] vs 184/2996 [6·1%]; RR 0·79, 0·64–0·97; p=0·03). 0607-1919. Treatment given more than 3 h after injury signifi cantly increased the risk of death due to bleeding (144/3272 Role of the funding source [4·4%] vs 103/3362 [3·1%]; RR 1·44, 1·12–1·84; p=0·004). The sponsor of the study had no role in study design, We recorded strong evidence that the eﬀ ect of data collection, data analysis, data interpretation, or tranexamic acid on death due to bleeding varied writing of the report. The corresponding author (IR) had according to time from injury to treatment (p<0·0001). full access to all the data in the study and had fi nal The evidence for interaction remained strong even after responsibility for the decision to submit for publication. adjustment for interactions between the other prespecifi ed baseline characteristics and treatment Results (p<0·0001; data not shown). Of the 3076 deaths from all causes, death due to bleeding The estimated OR of tranexamic acid on death due to accounted for 1063 (35%). The risk of death due to bleeding when given immediately after injury was 0·61 www.thelancet.com Published online March 24, 2011 DOI:10.1016/S0140-6736(11)60278-X 3
Take Home Messages • ABCDEに沿って評価していこう. • 急変したら, ABCDEに戻って評価しよう. • とにかくABCDEを安定化させよう.
JATEC Japan Advanced Trauma Evaluation and Care 2014年12月6-7日は広島コースです