Red blood cell to plasma ratios transfused during massive transfusion are associated with mortality in severe multiply injury: a retrospective analysis from the Trauma Registry of the Deutsche Gesellschaft für Unfallchirurgie
Department of Trauma and Orthopedic Surgery, and
Institute for Research in Operative Medicine (IFOM), University of Witten/Herdecke, Cologne-Merheim Medical Center (CMMC), Cologne, Germany
Conflict of interest statement: there are no conflicts of interest associated with this article.
: Marc Maegele, Department of Trauma and Orthopedic Surgery, University of Witten/Herdecke, Cologne-Merheim Medical Center (CMMC), Ostmerheimerstr. 200, D-51109 Cologne, Germany E-mail: firstname.lastname@example.org
Background To test whether an acute transfusion practice of packed red blood cells (pRBC) : fresh-frozen plasma (FFP) 1 : 1 would be associated with reduced mortality in acute bleeding multiply injury.
Methods Retrospective analysis using the TR-DGU database (Trauma Registry of the Deutsche Gesellschaft für Unfallchirurgie 2002–2006) on primary admissions with substantial injury (Injury Severity Score > 16) and massive transfusion (> 10 pRBCs). Seven hundred thirteen patients were divided into three groups according to the pRBC : FFP ratio transfused, that is, (i) pRBC : FFP > 1·1; (ii) pRBC : FFP 0·9–1·1 (1 : 1); and (iii) pRBC : FFP < 0·9, and mortality rates were compared.
Results Four hundred ninety-seven (69·7%) of patients were male, the mean age was 40·1 (± 18·3) years. Injury characteristics and pathophysiological state upon emergency room arrival were comparable between groups. Out of 713, 484 patients had undergone massive transfusion with pRBC : FFP > 1·1, 114 with pRBC : FFP 0·9–1·1 (1 : 1), and 115 with pRBC : FFP < 0·9 ratios. Acute mortality (< 6 h) rates for pRBC : FFP > 1·1, pRBC : FFP 0·9–1·1 (1 : 1), and pRBC : FFP < 0·9 ratios were 24·6, 9·6 and 3·5% (P < 0·0001), 24-h mortality rates were 32·6, 16·7 and 11·3% (P < 0·0001), and 30-day mortality rates were 45·5, 35·1 and 24·3% (P < 0·001). The frequency for septic complications and organ failure was higher in the pRBC : FFP 0·9–1·1 (1 : 1) group, ventilator days and length of stays for intensive care unit and overall in-hospital were highest in the pRBC : FFP < 0·9 ratio group (P < 0·0005).
Conclusions An association between pRBC : FFP transfusion ratios and mortality to favour early aggressive FFP administration was observed. Further investigation is necessary prior to recommending routine 1 : 1 or more aggressive FFP use in exsanguinating patients.
Uncontrolled haemorrhage is still a leading cause of early death in trauma  and the administration of blood products in the scenario of multiply injury is an essential and life-saving component of acute trauma care [2,3]. However, transfusion strategies for trauma-related haemorrhage are poorly defined and current recommendations are usually based upon experts’ opinions or personal experience rather than upon evidence from randomized-controlled trials. Recent military experience suggests that the early use of fresh-frozen plasma (FFP) together with packed red blood cells (pRBCs) in a 1 : 1 ratio for casualties requiring massive transfusion (> 10 pRBC units/24 h) is associated with improved outcome [1,4]. In reporting their transfusion experiences from the 2006 warfare at the Lebanon–Israeli border, Dann and colleagues  observed that early transfusion of FFP to casualties with penetrating wounds requiring massive transfusion is needed to overcome acute post-trauma coagulopathy. In addressing an early balanced transfusion therapy in massively bleeding patients, Johannson et al.  recently introduced an acute transfusion package (ATP) consisting of five pRBCs, five FFPs and two PCs (platelet concentration) units. A similar approach with a pre-defined ratio of the different blood products but based upon a validated guideline extended by timing of laboratory tests and a sound logistic protocol is urged by other authors [7,8]. More aggressive pre-ICU (intensive care unit) intervention to correct coagulopathy may be effective in decreasing pRBC requirements during ICU resuscitation, and, because of the association with increased mortality, could improve outcome . We reviewed datasets from multiply injured patients derived from the TR-DGU database (Trauma Registry of the Deutsche Gesellschaft für Unfallchirurgie/German Society of Trauma Surgery ) for massive transfusion practice between the years 2002 and 2006 to test the hypothesis whether an early aggressive transfusion practice of pRBC : FFP 1 : 1 between emergency room (ER) arrival and ICU admission would be associated with reduced mortality.
Materials and methods
We retrospectively analysed datasets from multiply injured patients that were entered into the TR-DGU database (German Society of Trauma Surgery ) between the years 2002 and 2006. TR-DGU was founded in 1993 but the use of FFP was not documented until the online version of the registry was introduced in 2002. Patients were eligible if the documentation comprised complete data on injury pattern and severity, pre- and in-hospital courses, relevant laboratory findings upon ER arrival, transfusion practice, that is, the use of pRBCs and FFP including ratios, and outcome. Only primary admissions and patients with substantial injury as reflected by an Injury Severity Score (ISS) > 16 who received a minimum of 10 pRBCs between ER arrival and ICU admission were considered. Patients that immediately died upon ER admission and never reached the ICU alive were excluded. Under these conditions, 713 patients were identified for further analysis.
The Trauma Registry of the Deutsche Gesellschaft für Unfallchirurgie (TR-DGU)
TR-DGU  was founded in 1993 by the German Society of Trauma Surgery and is run by a small steering group from different trauma centres in Germany (Working Group on Polytrauma/AG Polytrauma). Through 2006, data from a total of 29 353 trauma victims have been entered into the registry, with approximately 3000 new cases added each year. Since the introduction of the online version of the registry in 2002, the use of FFP unit is routinely documented. Between 2002 and 2006, 17 935 patients have been entered into the registry. Currently, there are 100 hospitals affiliated with the registry, mostly from Germany (n = 90), of which 70 are actually contributing data into the database. It is estimated that from the total number of severe trauma cases in Germany, approximately 30% are covered by the registry. It is not an obligatory registry and participation is free of charge. Data collection and documentation are structured in four consecutive time phases: A – the pre-hospital phase; B – emergency room and initial surgery (until admission to the ICU); C – ICU; and D – outcome status at discharge and description of injuries and procedures. The registry contains detailed information on demographics, injury pattern, co-morbidities, pre- and in-hospital management, time course, relevant laboratory findings including data on transfusion, and outcome of each individual. The trauma registry is approved by the review board of DGU and is in compliance with the institutional requirements.
Patients were divided into three groups according to the pRBC to FFP ratio transfused during immediate massive transfusion. Group 1 had undergone massive transfusion with a pRBC : FFP > 1·1, Group 2 had undergone massive transfusion with a pRBC : FFP 0·9–1·1 (1 : 1), and Group 3 had undergone massive transfusion with a pRBC : FFP < 0·9, and mortality rates (< 6-h, < 24-h, 30-day, and overall in-hospital mortality) were compared among groups. Data are presented as mean ± standard deviation (SD) for continuous variables and as percentages for incidence rates. All data were analysed by standard statistical software (SPSS, Chicago, IL, USA). Clinical data were compared between the groups using the χ2-test for categorical variables and the U-test for continuous variables. A P-value < 0·05 was considered significant.
A total of 713 multiply injured patients from the registry were identified for analysis. Four hundred ninety-seven (69·7%) patients were male and the mean age was 40·1 (± 18·3) years. Of 713, 484 patients had undergone massive transfusion with a pRBC : FFP > 1·1, 114 patients had undergone massive transfusion with a pRBC : FFP 0·9–1·1 (1 : 1), and 115 patients had undergone massive transfusion with a pRBC : FFP < 0·9 (Fig. 1). The number of pRBC and FFP units actually transfused as well as the basic characteristics of the patients within their respective study group are summarized in Table 1. There was no statistical difference between the three study groups with respect to injury characteristics, pathophysiologic state at scene and upon ER arrival, laboratory findings, frequency of emergency interventions and procedures and volume loading. Figure 2 depicts the rates of < 6-h, < 24-h and 30-day mortality for the three study groups. The rate for overall in-hospital mortality corresponded to that observed for 30-day mortality and is shown in Table 1. All rates were highest among patients that had undergone massive transfusion with a pRBC : FFP > 1·1. Of 484, 158 (32·6%) patients with this transfusion regimen died with the first 24 h after ER admission; the 30-day mortality rate for this group was 45·5% (220/484 patients). In contrast, acute (< 24-h) and 30-day mortality rates were lower in patients that were transfused with a pRBC : FFP 0·9–1·1 (1 : 1). In this group, 16·7% (19/114) of patients died within the first 24 h (P < 0·0001) with a 30-day mortality rate of 35·1% (40/115; P < 0·001). Most surprisingly, acute (< 24-h) and 30-day mortality rates were lowest in patients that had undergone massive transfusion with a pRBC : FFP < 0·9 ratio. Acute (< 24-h) and 30-day mortality rates for this group were 11·3% (13/115 patients; P < 0·0001) and 24·3% (28/115 patients; P < 0·001), respectively (Fig. 2). The rates for < 6-h mortality for the three study groups were 24·6% (119/484 patients), 9·6% (11/114 patients) and 3·5% (4/115 patients; P < 0·0001). The frequency for septic complications (non-significant) and single/multiorgan failure was higher in the pRBC : FFP 0·9–1·1 (1 : 1) group compared to the other two groups (P < 0·05); ventilator days and the length of ICU and overall in-hospital stays were highest in the group that was transfused with a pRBC : FFP < 0·9 (P < 0·005) (Table 1).
Table 1. Basic characteristics, injury pattern, physiologic state at scene and upon emergency room arrival and outcome of patients within their respective study groups (n = 713)
pRBC : FFP > 1·1
pRBC : FFP 0·9–1·1
pRBC : FFP < 0·9
1 unit = 230–260 ml.
1 unit = 220–280 ml.
P < 0.05;
P < 0.005.
AIS, abbreviated injury scale; BP, blood pressure; ER, emergency room; ICU, intensive care unit; IV, intravenous; LOS, length of stay; NS, non-significant; PPT, partial thromboplastin time; SD, standard deviation; WBC, white blood cells.
The present study was conducted to test the hypothesis whether an aggressive transfusion practice of pRBC : FFP 1 : 1 between ER and ICU would be associated with reduced mortality in severely injured patients requiring massive transfusion. Seven hundred thirteen multiply injured patients from the TR-DGU database with an ISS > 16 that had undergone massive transfusion with a minimum of 10 pRBCs were analysed. Comparable conditions given between all groups with respect to injury characteristics and pathophysiology at scene and upon ER arrival, mortality rates for acute and long-term mortality were consistently lower in the pRBC : FFP 0·9–1·1 (1 : 1) group vs. the pRBC : FFP > 1·1 group. However, the most surprising result from the present study was that mortality rates for all time points were lowest in the group that had undergone massive transfusion with a pRBC : FFP < 0·9.
Early aggressive correction of coagulopathy including immediate FFP substitution has been proven beneficial for survival following massive transfusion in patients that have undergone trauma . Recently, Gonzalez and colleagues  reported that resuscitation using a standard university massive transfusion protocol resulted in perpetuation of coagulopathy, but resuscitation with a pRBC : FFP 1 : 1 ratio resulted in correction of coagulopathy. While contributing factors such as acidosis and hypothermia are usually well-managed, coagulopathy may not be corrected even in the ICU despite adherence to pre-ICU mass transfusion and ICU protocols, likely due to inadequate or insufficient pre-ICU intervention [1,8]. Meanwhile, several experts have revised their pre-ICU transfusion protocol now emphasizing the early use of FFP in a pRBC:FFP 1 : 1 ratio [1,4]. Similarly, Dann et al.  in reporting their transfusion experiences from the 2006 warfare at the Lebanon–Israeli border observed that early transfusion of FFP to casualties with penetrating wounds requiring massive transfusion is needed to overcome acute post-trauma coagulopathy. By using a proactive approach for an early balanced transfusion therapy in massively bleeding patients, Johannson and colleagues  introduced an ATP consisting of five pRBCs, five FFPs and two PCs units. A similar approach with a pre-defined ratio of the different blood products but based upon a validated guideline extended by timing of laboratory tests and a sound logistic protocol is urged by other authors [7,8].
More aggressive pre-ICU intervention to correct coagulopathy, for example, by the early use of FFP in a fixed pRBC : FFP ratio may be effective in decreasing pRBC requirements during ICU resuscitation, and, because of the association with increased mortality, could improve outcome . Numerous studies have identified pRBC transfusions as an independent predictor for mortality [11–16], ICU admission [12,13,17], and lengths of ICU and in-hospital stays [11–13,17]. The transfusion of pRBCs has also been associated with the development of infectious complications [11,13,17–19], acute respiratory distress syndrome (ARDS) [14–16], and increased resource utilization ; all this in a dose-dependent fashion [15,18–20]. In the elderly, age and packed cell transfusion volume act independently, but yet synergistically to increase mortality following injury . The precise mechanism of pRBC transfusion-related complications with reduced outcome is yet to be established. Potential mechanisms include the generation of bioreactive lipids from pRBCs that have polymorphonuclear cell priming capacity that activate a systemic inflammatory response [22,23] or the activation of inflammatory genes in circulating leucocytes .
Recently, Schols et al.  investigated the result of therapeutic FFP transfusion to patients with dilutional coagulopathy on haemostatic activity. For all patients, FFP transfusion led to higher plasma coagulation factor levels, a shortened activated partial thromboplastin time, and a significant increase in thrombin generation. In this study, thrombin generation parameters and fibrinogen levels were higher in post-transfusion plasmas from patients who stopped bleeding compared to those with ongoing bleeding. Although the overall risk of FFP transfusion is low, together with platelets FFP are the least safe blood components due to potential immunologic reactions, transfusion-related acute lung injury (TRALI) and haemolysis resulting from anti-A or -B if transfused across ABO groups . In recent years, TRALI has developed from an almost unknown transfusion reaction to the most common cause of transfusion-related morbidities and fatalities [27,28]. TRALI is characterized by acute respiratory distress and non-cardiogenic lung oedema with temporal association to transfusion. TRALI is carefully to be differentiated from transfusion-related circulatory overload. In its acute presentation, TRALI can clinically be distinguished from acute respiratory distress occurring as a result of other causes. The pathophysiology of TRALI is still controversial. The extent of TRALI depends on the susceptibility of the patient to develop a more clinically significant reaction resulting from an underlying disease process and on the nature of potential triggers in the transfused blood components, including granulocyte-binding alloantibodies or neutrophil-priming substances . Immune TRALI, which occurs predominantly after the transfusion of FFP and platelet concentrates is rare (about 1 incidence per 5000 transfusions) but frequently (approximately in 70% of cases) requires mechanical ventilation and is not seldom fatal (6–9% of cases). In the present study, the frequency for septic complications and single/multiorgan failure was higher in the pRBC : FFP 0·9–1·1 (1 : 1) group and ventilator days as well as ICU and in-hospital stays were longest in the group that was transfused with a pRBC : FFP < 0·9 ratio.
Meanwhile, pathogen reduction for FFP is well-established in Europe. Furthermore, a policy eliminating the use of transfusible plasma from female donors that was initiated in the UK in 2003 has been credited with a reduced incidence of TRALI fatalities from 23 cases in 2003, to six cases in 2005, and no plasma-associated TRALI deaths in 2005 . In the USA, a retrospective review of 38 TRALI-associated fatalities has shown that 24 (63%) followed plasma transfusion. Female donors positive for allogeneic white blood cell antibodies were involved in 75% of these plasma cases and in 71% of all TRALI fatalities .
The present study has several limitations. First, it is a retrospective analysis limited by the data that were available from the TR-DGU database with only those patients included from which complete datasets existed. In this context, the authors are aware of the fact that registry data like prospective observational studies share similar limitations with respect to comparability of therapeutic subgroups. Therefore, observed associations do not necessarily represent causality. Second, detailed data on the development of coagulopathy and infectious complications within the individual's further sequelae on ICU after having been exposed to different pRBC : FFP transfusion ratios were not completely available. Third, the potential impact of the obviously unequal distribution of patients into the different transfusion groups needs to be considered. In the present study, 67·9% of the patients had received transfusions in a pRBC : FFP > 1·1, 15·9% in a pRBC : FFP 0·9–1·1 (1 : 1), and 16·2% in a pRBC : FFP < 0·9 ratio. The higher number of patients that was transfused with a pRBC : FFP > 1·1 ratio possibly reflects the common attitude in the past to transfuse more pRBCs per se in the multiply injured patient.
Nevertheless, in the present study immediate pRBC : FFP 0·9–1·1 (1 : 1) transfusion was associated with enhanced survival in severe multiply injured patients requiring massive transfusion. Given the result in that lowest mortality rates were observed in those patients that had undergone massive transfusion with higher amounts of FFP compared to pRBCs during initial care, one may suggest an even more aggressive and proactive approach towards the correction of acute post-traumatic coagulopathy. Potential strategies to quickly restore haemostasis may include not only FFP but also other components [31–34].
The present findings suggest an association of pRBC : FFP transfusion ratios with mortality in favour of early and aggressive FFP administration, but further clinical investigation is necessary prior to recommending routine 1 : 1 or more aggressive FFP use in the exsanguinating trauma patient.
Participating departments and hospitals of the German Trauma Registry (DGU-Traumaregister) in alphabetical order of the city:
Unfallchirurgie der Universität Aachen, Klinik für Unfall- und Wiederherstellungschirurgie des Zentralklinikums Augsburg, Unfallchirurgie und Orthopädie der Klinik Bad Hersfeld, Klinik für Unfall- und Wiederherstellungschirurgie der Charite – Campus Virchow-Klinikum Berlin, Unfallchirurgie des Martin – Luther Krankenhauses Berlin, Chirurgische Klinik des Klinikums Berlin–Buch, Berufsgenossenschaftliche Unfallklinik des Krankenhauses Berlin-Mahrzahn, Unfallchirurgische Klinik der Krankenanstalten Gilead Bielefeld, Chirurgische Klinik und Poliklinik BG-Klinik Bochum Bergmannsheil, Unfallchirurgische Klinik des Knappschaftskrankenhauses der Ruhr-Universität Bochum, Klinik für Unfallchirurgie der Rheinischen Friedrich-Wilhelms-Universität Bonn, Klinik für Unfall- und Wiederherstellungschirurgie des Zentralkrankenhauses Sankt-Jürgen-Straße Bremen, Zentrum für Allgemein- und Unfallchirurgie am Zentralkrankenhauses Ost Bremen, Unfallchirurgische Klinik des Zentralkrankenhauses Reinekenheide Bremerhaven, Unfallchirurgie des Kreiskrankenhauses Burg, Unfallchirurgische Abteilung des Allgemeinen Krankenhauses Celle, Klinik für Unfall- und Gelenkchirurgie des Klinikum Chemnitz, Klinik für Unfall- und Handchirurgie Dessau, Unfall- und Wiederherstellungschirurgie am Klinikum Lippe-Detmold, Abteilung für Chirurgie des Krankenhauses Dresden-Neustadt, Klinik für Unfall- und Wiederherstellungschirurgie der Technischen Universität Dresden, Klinik für Unfallchirurgie des Städtischen Klinikums Friedrichstadt, Klinik für Allgemein- und Unfallchirurgie der Heinrich-Heine-Universität Düsseldorf, Unfall-, Hand- und Wiederherstellungschirurgie des Klinikums Erfurt, Unfallchirurgie des Kreiskrankenhauses Eschwege, Unfallchirurgische Klinik der Universität Essen, Unfallchirurgische Abteilung des Evangelischen Krankenhauses Lutherhaus Essen, Berufsgenossenschaftliche-Unfallklinik Frankfurt/Main, Zentrum für Chirurgie Klinik für Unfallchirurgie der Universität Frankfurt/Main, Chirurgische Klinik des Klinikums Frankfurt/Oder, Unfallchirurgie, orthopädische und wiederherstellende Chirurgie des Klinikums Fürth, Chirurgie des Johanniter-Krankenhauses Geesthacht, Unfall- und Wiederherstellungschirurgie des Städtischen Klinikums Görlitz, Klinik für Anästhesiologie, Operative Intensivmedizin und Schmerztherapie der Klinik am Eichert Göppingen, Zentrum für Chirurgie der Georg-August-Universität Göttingen, Unfallchirurgie des Evangelischen Krankenhauses Göttingen-Weende, Universitätsklinik für Unfallchirurgie Graz, Unfallkrankenhaus Graz der Allgemeinen Unfallversicherungsanstalt Graz, Chirurgie des Kreiskrankenhauses Grevenbroich, Klinik für Traumatologie der Universität Groningen, Unfallchirurgische Abteilung Krankenhaus Gummersbach, Unfall- und Wiederherstellungschirurgie am Berufsgenossenschaftliches-Unfallkrankenhaus Hamburg, Abteilung für Unfall- und Wiederherstellungschirurgie der Universitätsklinik Hamburg-Eppendorf, Abteilung Unfallchirurgie Kreiskrankenhaus Hameln, Unfallchirurgische Abteilung des Städtischen Krankenhauses Hannover Nordstadt, Unfallchirurgische Abteilung des Friederikenstifts Hannover, Unfallchirurgische Klinik der Medizinischen Hochschule Hannover, Chirurgische Abteilung Krankenhaus Hattingen, Abteilung Orthopädie I der Orthopädischen Universitätsklinik Heidelberg, Klinik für Unfall- und Wiederherstellungschirurgie des St. Bernward Krankenhauses Hildesheim, Unfallabteilung des Waldviertel Klinikums Horn, Abteilung für Unfallchirurgie des LKH Judenburg-Knittelfeld, Unfallchirurgische Klinik der Universität Homburg-Saar, Unfall-, Hand- und Wiederherstellungschirurgie des Klinikums Karlsruhe, Klinik für Unfallchirurgie der Universität Kiel, Unfallchirurgische Klinik I. Chirurgischer Lehrstuhl der Universität zu Köln, Lehrstuhl für Unfallchirurgie/Orthopädie der Universität Witten/Herdecke am Klinikum Köln-Merheim, Abteilung für Unfallchirurgie des Allgemeinen Öffentlichen Krankenhauses Krems/Donau, Unfall- und Wiederherstellungschirurgie des Städtischen Klinikums St. Georg Leipzig, Klinik für Unfall- und Wiederherstellungschirurgie des Universitätsklinikums Leipzig, Chirurgische Abteilung des Evangelischen Krankenhauses Lengerich, Abteilung für Unfallchirurgie des Allgemeinen öffentlichen Krankenhauses Linz, Evangelisches Krankenhaus Lippstadt, Operatives Zentrum I Klinik für Chirurgie Universitätsklinikum Lübeck, Berufsgenossenschaftliche-Unfallklinik Ludwigshafen, Klinik für Unfall- und Wiederherstellungschirurgie des St. Marien-Hospital Lünen, Abteilung für Unfallchirurgie der Otto-v.-Guericke-Universität Magdeburg, Klinik für Chirurgie des Klinikums Magdeburg Krankenhaus Altstadt, Klinik für Unfallchirurgie der Universität Mainz, Abteilung für Unfallchirurgie des Universitätsklinikums Mannheim, Klinik für Unfallchirurgie der Philipps-Universität Marburg, Unfallchirurgie des Klinikums Minden, Unfallchirurgie des Krankenhauses Maria Hilf Mönchengladbach, Chirurgische Klinik Klinikum Großhadern der Ludwig Maximilian Universität München, Unfallchirurgische Abteilung Klinikum München-Harlaching, Chirurgische Klinik Klinikum Innenstadt der Ludwig Maximilian Universität München, Klinik für Unfall- und Handchirurgie der Westfälischen Wilhelms-Universität Münster, Berufsgenossenschaftliche-Unfallklinik Murnau, Klinik für Unfall- und Wiederherstellungschirurgie der Städtischen Kliniken Lukaskrankenhaus Neuss, Unfallchirurgie des Marienhospitals Osnabrück, Unfallchirurgie des Vogtland Klinikums Plauen, Unfallchirurgie des Krankenhauses Lennep Klinikum Remscheid, Klinik für Unfall- und Wiederherstellungschirurgie Klinikum Rosenheim, Chirurgie und Unfallchirurgie Sana-Krankenhaus Rügen, Landesklinik für Unfallchirurgie und Sporttraumatologie des Landeskrankenhauses Salzburg, Unfallchirurgie Diakoniekrankenhaus Schwäbisch-Hall, Chirurgische Klinik des Kreiskrankenhauses Soltau, Unfallchirurgie im Johanniter-Krankenhaus der Altmark in Stendal, Unfall- und Wiederherstellungschirurgie des Kreiskrankenhauses Traunstein, Berufsgenossenschaftliche-Unfallklinik Tübingen, Unfallchirurgie des Bundeswehrkrankenhauses Ulm, Abteilung für Unfall-, Hand- und Wiederherstellungschirurgie der Universität Ulm, Unfallchirurgische Klinik des Klinikums Villingen-Schwenningen, Unfallchirurgie Klinikum Weiden, Unfallchirurgie des Asklepios Krankenhauses Weißenfels, Abteilung für Unfallchirurgie und Sporttraumatologie des Donauspitals Wien, Unfallchirurgische Klinik der Universität Würzburg, Klinik für Unfall- und Wiederherstellungschirurgie des Klinikums Wuppertal, Unfallchirurgische Klinik der Universität Zürich, Rettungswache Zusmarshausen.