Background
Methods
Information sources and search technique
Eligibility criteria and study selection
Data abstraction and analysis
Results
Study characteristics
Author, region | Study design, centers, patients, years |
N
| Objective | ISS (mean ± SD or median/IQR) | Age (mean or median) | Sex male, n (%) |
---|---|---|---|---|---|---|
Rugeri, 2006 [36] France | Single center prospective, civilian, Jul 2004-Oct 2004 | 88 | Detect coagulopathy Guide transfusion | 22 (12–34) | 34 (±16) | 68 (77 %) |
Levrat, 2008 [37] France | Single center prospective, civilian, Jul 2004-Oct 2004 | 87 | Diagnosis of HF | Control group: 20 (11–29) HF group: 75 (75–75) | Control: 29 (21–43) HF: 30 (24–45) | Control: 64/82 (78 %) HF: 4/5 (80 %) |
Schochl, 2009 [26] Austria | Single center prospective, civilian, Jan 2003-Dec 2007 | 33 | Diagnosis of HF Predict mortality | 47 ± 14 | 45 (20–88) | 22 (67 %) |
Doran, 2010 [25] Afghanistan | Single center prospective, military, Jan 2009-Mar 2009 | 25 | Detect coagulopathy | MT group: 35 (25–50) Non MT: 20 (19–20) | 21 (18–35) | 25 (100 %) |
Leemann, 2010 [38] Zurich | Single center retrospective, civilian, Jan 2006-Dec 2006 | 53 | Predict MT | 31.1 ± 1.7 | 39.6 (±2.5) | 40 (75.5 %) |
Schochl, 2010 [27] Austria | Single center retrospective, civilian, Jan 2005-Apr 2009 | 131 | Guide transfusion | 38 ± 15 | 46 ± 18 | 96 (73 %) |
Tauber, 2011 [40] Austria | Single center, prospective, civilian, Jul 2005-Jul 2008 | 334 | Detect coagulopathy Predict RBC transfusion Predict mortality | 34 (24–45) | 43 (27–56) | 260 (77.8 %) |
Schochl, 2011 [39] Austria | Single center, retrospective, civilian, Jan 2005-Oct 2010 | 88 | Predict mortality | Survivors: 20 (16–26.25) Non survivors 29 (25–30.75) | 47 (26–66) | 67 (76 %) |
Schochl, 2011 [28] Austria | Single center retrospective, civilian, Jan 2005-Dec 2010 | 323 | Predict mortality | Non-MT group: 27 (20–34) MT group: 42 (34–50) | 44 (26–59) | 245 (78 %) |
Davenport, 2011 [41] United Kingdom | Single center prospective, civilian, Jan 2007-Jun 2009 | 300 | Detect coagulopathy Predict MT | 12 (4–25) | 33 (23–48) | 246 (82 %) |
Rourke, 2012 [42] England | Multicenter prospective, civilian, Jan 2008-Dec 2010 | 517 | Detect coagulopathy Guide transfusion | 14 (8–27) | 36 (23–51) | 405 (78 %) |
Woolley, 2013 [30] Afghanistan | Single center prospective, military, May 2009-Jul 2009 | 48 | Predict coagulopathy | 34 (17–43) | 24 (21–26) | 48 (100 %) |
Hagemo 2015 [43] | Multi center prospective civilian, Jan 2007-Nov2011 | 808 | Detect coagulopathy, Predict MT | 16 (20) | 38 (28) | 625 (77.4 %) |
Methodological quality
Reference | Representativeness of the exposed cohort | Selection of non-exposed cohort | Ascertainment of exposure | Outcome not present at start | Comparability of controls | Assessment of outcome | Adequate follow up | Loss to follow up | Total score |
---|---|---|---|---|---|---|---|---|---|
Rugeri 2007 [36] | * | - | * | * | - | * | * | * | 6/9 |
Levrat 2008 [37] | * | - | * | * | - | * | * | * | 6/9 |
Schöchl 2009 [26] | * | - | * | * | - | * | * | * | 6/9 |
Doran 2010 [25] | * | - | * | * | - | * | * | - | 5/9 |
Leemann 2010 [38] | * | - | * | * | - | * | * | * | 6/9 |
Schochl 2010 [27] | * | - | * | * | - | * | * | * | 6/9 |
Tauber 2011 [40] | * | - | * | * | - | * | * | * | 6/9 |
Schochl 2010 [39] | * | - | * | * | - | * | * | * | 6/9 |
Davenport 2011 [41] | * | - | * | * | - | * | * | * | 6/9 |
Schöchl 2011 [28] | * | - | * | * | - | * | * | * | 6/9 |
Rourke 2012 [42] | * | - | * | * | - | * | * | * | 6/9 |
Woolley 2012 [30] | * | - | * | * | - | * | * | * | 6/9 |
Hagemo 2015 [43] | * | - | * | * | - | * | * | * | 6/9 |
Risk of bias | Applicability concerns | ||||||
---|---|---|---|---|---|---|---|
Reference | Patient selection | Index test | Reference standard | Flow and timing | Patient selection | Index test | Reference standard |
Rugeri 2007 [36] | ☺ | ☹ | ☺ | ☺ | ☺ | ☹ | ☺ |
Levrat 2008 [37] | ☺ | ? | ☹ | ☺ | ☺ | ☺ | ☹ |
Schöchl 2009 [26] | ☺ | ? | ? | ☺ | ☺ | ☺ | ☺ |
Doran 2010 [25] | ☹ | ? | ? | ☺ | ☹ | ☺ | ☺ |
Leemann 2010 [38] | ☺ | ? | ? | ☺ | ☺ | ☺ | ☺ |
Schochl 2010 [27] | ☹ | ☺ | ☺ | ☺ | ☺ | ☺ | ☺ |
Schochl 2011 [39] | ☺ | ? | ? | ☺ | ☺ | ☺ | ☺ |
Davenport 2011 [41] | ☺ | ? | ? | ☺ | ☺ | ☺ | ☺ |
Schöchl 2011 [28] | ☺ | ? | ? | ☺ | ☺ | ☺ | ☺ |
Tauber 2011 [40] | ☺ | ☺ | ☺ | ☺ | ☹ | ☺ | ☺ |
Rourke 2012 [42] | ☺ | ? | ? | ☺ | ☺ | ☺ | ☺ |
Woolley 2012 [30] | ☺ | ? | ☺ | ☺ | ☺ | ☺ | ☺ |
Hagemo 2015 [43] | ☹ | ? | ? | ☺ | ☺ | ☺ | ☺ |
Outcomes
Studies addressing thresholds of ROTEM® parameters to diagnose ACoTS
Study | Comparator | ROTEM® thresholds used | Accuracy of threshold Sensitivity/Specificity | AUC | Key findings | |
---|---|---|---|---|---|---|
Rugeri 2006 [36] | PTR > 1.5 FIB < 1.0 g/L | EXTEM CA15 = 32 mm FIBTEM A10 = 5 mm | 87 91 | 100 85 | 0.98 0.96 | 1 – Significant correlation between EXTEM CA15 < 32 mm and PT >1.5 (r = 0.66, p < 0.0001) and of FIBTEM CA10 < 5 mm and Fibrinogen <1.0 g/L (r = 0.85, p < 0.0001). 2 – EXTEM A15 = 32 and FIBTEM A10 = 5 mm had a higher sensitivity and specificity to detect PTr > 1.5 and fibrinogen <1.0 g/L |
Levrat 2008 [37] | ELT < 90s | EXTEM MCF ≤ 18 mm LI30 ≤ 71 % APTEM MCF ↑by 7 % | 100 75 80 | 100 100 100 | 1.00 0.87 0.92 | 1 – MCF correlated well with ELT when compared with amplitude and CLI. 2 – HF patients exhibited greater ROTEM® abnormalities, lower INR, lower fibrinogen levels and were more severely injured (↑ ISS) compared to the control group (all p < 0.05) |
Schochl 2009 [26] | No comparator | EXTEM and INTEM ML = 100 % | NA | NA | NA | 1 – Fulminant HF confirmed by complete clot lysis within 30 min by ROTEM® trace |
Doran 2010 [25] | PT > 18 s aPTT > 38 s | EXTEM MCF < 45 mm | NA | NA | NA | 1 – ROTEM® detected coagulation abnormalities in 64 % patients vs. 10 % detected by SCTs as compared to test reference ranges? (p = 0.0005). 2 – MCF < 45 mm in 100 % of MT patients |
Davenport 2011 [41] | PTR > 1.2 | EXTEM CA5 ≤ 35 mm | 77 | NA | NA | 1 – EXTEM CA5 ≤ 35 mm detected coagulopathy with 77 % sensitivity and a false positive rate of 13 % |
Tauber, 2011 [40] | INR > 1.5 aPTT > 50s | EXTEM MCF < 45 mm | 72 | 76 | 0.83 | 1 - Prevalence of low fibrinogen, impaired fibrin polymerization and reduced MCF was 26 %, 30 %, and 22 %, respectively, higher than the prolonged INR (14 %) 2 – There was ↑ F1 + 2 and TAT and low AT levels, indicating ↑ thrombin formation among all patients |
FIB < 1.5 g/L PLT < 100 × 103
| FIBTEM MCF < 7 mm LI60 < 85 % | 86 79 | 71 78 | 0.89 0.84 | ||
Schochl, 2011 [39] | PTI < 70 %, | EXTEM CT > 80s CFT > 159 s MCF < 50 mm | NA | NA | 0.77 | 1 – Coagulopathy was characterized by abnormal values in most or all ROTEM® measurements as compared to reference range vs. SCT. 2 – Significantly low CA5-CA30 min, MCF in EXTEM, INTEM and FIBTEM assays in non survivors vs. survivors (p < 0.01) |
aPTT > 35 s, PLT < 100 × 103
| INTEM CT > 240 s CFT > 110 s MCF < 50 mm | |||||
FIB < 1.5 g/L | FIBTEM MCF < 9 mm | |||||
Rourke 2012 [42] | FIB < 1.5 g/L | EXTEM CA5 < 36 mm FIBTEM CA5 < 9.5 mm | 53 78 | 87 70 | NA NA | 1 – ROTEM® parameters correlated with fibrinogen level. 2 – Ex vivo fibrinogen administration reversed coagulopathy by ROTEM®. |
Woolley 2012 [30] | PT > 1.5 | EXTEM CA5 < 32 mm EXTEM A10 < 40 mm | 96 100 | 58 70 | NA NA | 1 – EXTEM MCF < 40 mm and interim values of EXTEM A5 and A10 predicted coagulopathy (A15: sensitivity/specificity of 96 %/58 % and for A10: sensitivity/specificity 100 %/70 %) |
Hagemo 2015 [43] | INR > 1.2 | EXTEM CA5 < 37 mm FIBTEM CA5 < 8 mm | NA NA | NA NA | 0.79 0.80 | 1 – Highest ROTEM® AUC values were found for EXTEM CA5 and FIBTEM CA5 for detecting ACoTS 2 – EXTEM CA5 ≤ 37 mm had a detection rate of 66.3 % and FIBTEM CA ≤ 8 mm had a detection rate of 67.5 % of ACoTS |
ROTEM® thresholds determined with comparison to standard controls (SCTs)
Hypofibrinogenemia
Hyperfibrinolysis
Platelet count
Studies addressing thresholds of ROTEM® parameters in predicting or guiding transfusion
Predicting transfusion
Study | Comparator | ROTEM® thresholds used | Accuracy of threshold Sensitivity/Specificity | ROC/AUC | Key findings | |
---|---|---|---|---|---|---|
Massive transfusion | ||||||
Leemann 2010 [38] | aPTT > 36 s PLT < 100 × 103
INR > 1.2 | EXTEM/INTEM CA10, CA20, CFT, MCF as per manufacturer | NA | NA | 0.82 | 2 – INTEM MCF 37.5 ± 2.9 associated with MT requirements within 24 h |
Tauber, 2011 [40] | FIB 1.50 g/L INR > 1.5 | FIBTEM MCF < 7 mm | 71 | NA | 0.80 | 1 – FIBTEM MCF < 7 mm associated with RBC use (OR 0.92, 95 % CI 0.87–0.98) |
Schochl 2011 [28] | PLT ≤ 161 × 103
aPTT ≤ 35.2 s FIB ≤ 1.4 g/dL | FIBTEM A10 ≤ 4 mm FIBTEM MCF ≤ 7 mm | 63.3 77.5 | 83.2 74.9 | 0.83 0.84 | 1 – 85 % patients with FIBTEM MCF 0–3 mm received MT 2 – FIBTEM A10 (0.83) and FIBTEM MCF (0.84) showed high predictive value for MT |
Davenport 2011 [41] | PTR > 1.2 | EXTEM CA5 ≤ 35 mm | 71.4 % | NA | NA | 1 – CA5 identified patients who required MT with detection rate of 71 % vs. 43 % for PTR > 1.2, p < 0.001 |
Hagemo 2015 [43] | INR > 1.2 | EXTEM CA5 ≤ 40 mm FIBTEM CA5 ≤ 9 mm | 72.7 % | 77.5 % | 0.75 0.78 | 1 – ROTEM CA5 is a valid predictor for MT. |
Any transfusion | ||||||
Schochl, 2010 [27] | PT (11–13.5 s) aPTT (26-35 s) FIB (2–4.5 g/L) PLT (150–350) | FIBTEM MCF < 10 mm EXTEM CT > 1.5× normal | NA | NA | NA | 1 – ROTEM® guided FC and PCC transfusion, associated with favorable survival (24.4 % vs. 33.7 %; p = 0.032) |
Davenport 2011 [41] | PTR > 1.2 | EXTEM CA5 ≤ 35 mm CT > 94 s Alpha < 650
| 33.3 % | NA | NA | 1 – CA5 ≤ 35 mm predicted RBC and plasma transfusion. Patients with CA5 ≤ 35 mm received more RBC (46 % vs. 17 %, p < 0.001) and plasma (37 % vs. 11 %, p < 0.001) transfusions. 2 – CA5 ≤ 35 mm received more RBC (4U vs. 1U, p < 0.001) and FFP (2U vs. 0U, p < 0.001) |
Guiding transfusion
Studies addressing thresholds of ROTEM® parameters in predicting mortality
Study | Comparator | Optimal ROTEM®® Parameter and cut off | Accuracy of threshold | Key findings | ||
---|---|---|---|---|---|---|
Sensitivity | Specificity | AUC | ||||
Levrat 2008 [37] | ELT < 90 min | EXTEM MCF ≤ 18 mm LI30 ≤ 71 % APTEM MCF ↑ by 7 % | 100 75 80 | 100 100 100 | 1.00 0.87 0.80 | 1 – Patients with HF had higher mortality rate (100 %, CI: 48–100 % vs. 11 % CI: 5–20 %, p < 0.05) |
Schochl 2009 [26] | No comparator | ML = 100 % | NA | NA | NA | 1 – Fulminant HF associated with 100 % mortality 2 – ↑CFT and ↓PLT contribution to MCF associated with ↑mortality (p = 0.042 and p = 0.026 respectively) |
Schochl 2010 [27] | No comparator | FIBTEM MCF < 10 mm EXTEM CT > 1.5 × normal | NA | NA | NA | 1 – Observed mortality was lower than the predicted mortality by TRISS (24.4 % vs.33.7 %, p = 0.032) with a favourable survival rate. |
Tauber, 2011 [40] | PT = 70 % FIB = 1.82 g/L | FIBTEM MCF < 7 mm, EXTEM CT 91 s EXTEM CFT 218 s EXTEM MCF 46 mm | NA | NA | 0.8 | 1 – FIBTEM MCF < 7 mm and EXTEM MCF < 45 mm associated with higher mortality (21 % vs. 9 % SCTs, p = 0.006 and 25.4 % vs. 9.4 % SCTs, p < 0.001, respectively) 2 – EXTEM MCF had strong association with early deaths (OR 0.94, 95 % CI 0.9–0.99). |
Schochl, 2011 [39] | aPTT > 35 s | FIBTEM MCF < 9 mm | NA | NA | 0.77 | 1 – Decrease in clotting times in EXTEM and INTEM (p < 0.001), decreased CFT in EXTEM and INTEM (p < 0.0001), and increased MCF in EXTEM, INTEM, and FIBTEM (p < 0.01) were noted in survivors compared with non-survivors, in patients with severe isolated TBI 2 – FIBTEM MCF (ROC 0.77, 95 % CI 0.66.5–0.85, p < 0.001) and aPTT (ROC 0.79 95 % CI 0.68–0.86, p < 0.001) independently associated with mortality. |
Rourke, 2012 [42] | FIB < 1.5 g/L | EXTEM CA5 < 36 mm FIBTEM CA5 < 10 mm | 53 78 | 87 70 | NA NA | 1 – Fibrinogen level was independently associated with higher mortality at 24 h and 28 days (p < 0.001). ROTEM could detect hypofibrinogenemia early and rapid replacement of fibrinogen may improve outcomes. |