Abstract

Introduction

Many advances have been achieved in the care of the polytrauma patient, leading to an improvement in survival and hospitalization rates [1]. In addition, the definition of polytrauma has been refined. Injury severity score (ISS) or related scoring system such as the Hospital Trauma Index (HTI) and later also the New Injury Severity Score (NISS), which are injury-specific classifications, were refined by adding patient specific factors or the physiological responses [2]. In contrast, the term “major fracture” is commonly used but yet to be sufficiently defined [3]. High-energy trauma is commonly associated with extremity fractures and severe tissue lesions. Osseous injuries represent the key focus for the orthopedic surgeon and the timing and technique of fixation is of pivotal importance for patient mobilization [4], pain management [4], the systemic inflammatory response [5] and the overall outcome [4]. Fractures that should be treated with increased priority have long been termed “major fractures” and most frequently, the term had been used for long bone injuries. Recently, the notion to find a universal consensus on the definition of a “major fracture” was introduced [3], motivated by an initiative of the polytrauma section of the European Society for Trauma and Emergency Surgery (ESTES) [6].

In addition, several terminology issues have been addressed in preparing the consensus process, which was initiated at ECTES in 2019, followed by further scientific sessions, in-person discussions and structured discussion groups during courses [3].

A recent survey of this international panel of experienced surgeons suggested that the anatomic location of a fracture should no longer be the only focus of attention. Instead, it has been indicated that a fracture should be considered a “major fracture”, when it drives the surgical treatment strategy [3, 6]. This, in turn, can be brought about by a number of reasons, ranging from physiological derailment [7] to relevant soft tissue injury [8]. These considerations have been supported by numerous examples in the literature and by the latest revision of the abbreviated injury scale: The 2015 version of AIS assigns fractures a higher score in case they are open, or accompanied by a vascular or neurological injury, emphasizing the role of injury-specific factors and the potential for a physiological response [5, 9].

Subsequently, it has been argued that improvements in the terminology may be helpful to develop optimal patient care and Safe Definitive Surgery (SDS) [10, 11]. As part of a consensus process of the polytrauma section of ESTES, we aim to prepare the ground for defining the “major fracture” by revisiting the literature.

This manuscript aims to investigate how fractures have influenced surgical decision-making in the scientific literature. We also intend to identify relevant factors (i.e., physiological, concomitant injuries) that determine treatment strategies and the management of a “major fracture”.

Methods

The reporting of this systematic review is in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [12]. We performed a systematic search to identify all relevant original publications that investigated surgical treatment strategy in the polytraumatized patient (operative timing/sequence, type of surgery and decision-making). To approximate how “major fracture” was defined, we assessed the prevalence of the anatomic regions of interest and the core determining factors for the surgical treatment strategy. A qualitative synthesis was performed.

Definitions

Chest injury was defined as a relevant injury to the chest that would correspond to an abbreviated injury scale (AIS) or HTI of at least two points. They include, i.e., clinical or radiological lung contusions, serial rib fractures or an initial worsening of respiratory parameters.

Traumatic brain injury (TBI) was defined as a relevant injury to the head that would correspond to an abbreviated injury scale (AIS) or HTI of at least two points.

Shock was defined as hemodynamic instability, measured either through abnormal vital signs (sBP<90 mmHg, HR>100), a derailed acid–base balance (lactate>2 mmol/l, BE<-4) or the need for blood transfusions.

“Other” relevant injury was defined as either the presence of a spinal cord injury (SCI) with a neurological deficit, or the presence of a relevant soft tissue injury including open fractures≥type two (Gustilo-Anderson), or the presence of a relevant abdominal injury, either requiring surgical intervention or at high risk for causing hemodynamic instability.

Results

Study selection

The flow chart of the study selection process is presented in Fig. 1: The systematic search in the Medline Database yielded 2900 results and 1329 records were identified through EMBASE. A further 49 records were identified through additional sources. After removal of 144 duplicates, 4134 records were screened and 4002 records were excluded. The remaining 132 articles were sought for retrieval. While 10 articles could not be retrieved, 122 full- text articles were assessed for eligibility. After exclusion of 48 articles, 74 studies remained for qualitative synthesis (complete article information/metadata are presented in the appendix) [4, 7, 14–85].

Open in a separate window

Fig. 1

Flowchart of the selection process

Overview

A synopsis of all included studies is presented in Table ​Table1.1. In total, 74 studies published between 1982 and 2022 matched our criteria and were included. The majority were published in the decades from 1990 to 1999 (20 studies) and from 2010 and 2019 (24 studies). Among the included studies, 50 (67.6%) focused on one single anatomic region, while the remaining 24 (32.4%) studies investigated multiple anatomic regions. Among the publications on one anatomic region, 40 (80%) studies focused on femur fractures, 5 (10%) studies on pelvic fractures and 5 (10%) studies on spinal fractures. Among the publications that addressed multiple regions, 22 (91.7%) included femur fractures and 17 (70.8%) included pelvic fractures. Further anatomic regions of interest were tibia, spine, acetabulum and the upper extremities (see Fig. 2 and Table ​Table22).

Table 1

Synopsis of general information, relevant anatomic regions and determining factors in all included articles, sorted by year of publication

General information			Relevant anatomic regions							Determining factors
Authors	Year	n=	Focus on	Femur	Pelvis	Acetabulum	Spine	Tibia	UEX	Chest Injury	TBI	Shock	Other
Goris et al. [14]	1982	58	Multiple regions	+				+	+
Browner et al. [15]	1984	54	One region only	+							+		+
Sturm et al. [16]	1984	207	One region only	+						+
Johnson et al. [17]	1985	132	Multiple regions	+	+		+			+
Seibel et al. [18]	1985	56	Multiple regions	+		+				+	+		+
Meek et al. [19]	1986	71	Multiple regions	+				+	+
Brug et al. [20]	1988	361	One region only	+						+
Bone et al. [4]	1989	83	One region only	+						+		+
Burchardi et al. [21]	1990	225	Multiple regions	+				+	+	+
Nast-Kolb et al. [22]	1990	69	One region only	+						+
Hofman et al. [23]	1991	58	Multiple regions	+	+		+	+			+
Pelias et al. [25]	1992	130	Multiple regions	+	+		+			+
Poole et al. [26]	1992	114	Multiple regions	+				+			+
Pape et al. [24]	1992	16	One region only	+						+
Riemer et al. [27]	1992	150	One region only	+									+
Pape et al. [28]	1993	106	One region only	+						+
Pape et al. [29]	1993	319	One region only	+						+
Bone et al. [30]	1994	676	Multiple regions	+	+	+		+			+
Malisano et al. [32]	1994	153	Multiple regions	+				+	+		+
van Os et al. [33]	1994	27	Multiple regions	+	+	+	+	+	+	+	+
Charash et al. [31]	1994	138	One region only	+						+
Reynolds et al. [34]	1995	105	One region only	+						+	+
Van Der Made et al. [35]	1996	60	One region only	+						+
Bosse et al. [36]	1997	453	One region only	+						+
Boulanger et al. [37]	1997	149	One region only	+						+
Schmidtmann et al. [38]	1997	17	One region only	+						+	+
Aufmkolk et al. [39]	1998	325	One region only	+						+
McLain et al. [40]	1999	26	One region only				+						+
Nowotarski et al. [41]	2000	54	One region only	+								+	+
Scalea et al. [42]	2000	327	One region only	+						+	+	+	+
Taeger et al. [45]	2002	45	Multiple regions	+	+			+	+	+	+	+	+
Brundage et al. [43]	2002	674	One region only	+						+	+
Pape et al. [44]	2002	514	One region only	+						+	+	+	+
Bhandari et al. [46]	2003	1211	Multiple regions	+				+			+
Nau et al. [47]	2003	352	One region only	+						+	+
Pape et al. [48]	2003	35	One region only	+						+
Taeger et al. [50]	2005	409	Multiple regions	+	+			+	+	+	+	+
Harwood et al. [49]	2005	174	One region only	+						+	+	+	+
Powell et al. [51]	2006	315	One region only	+									+
Pape et al. [52]	2007	165	One region only	+						+	+	+	+
Probst et al. [53]	2007	290	One region only		+
Morshed et al. [54]	2009	3069	One region only	+									+
O'Toole et al. [55]	2009	227	One region only	+						+
Tuttle et al. [56]	2009	462	One region only	+							+	+
Vallier et al. [60]	2010	418	Multiple regions		+	+				+		+
Hartsock et al. [58]	2010	19	One region only	+						+
Scannell et al. [59]	2010	205	One region only	+						+
Enninghorst et al. [57]	2010	45	One region only		+							+
Schreiber et al. [63]	2011	114	Multiple regions	+	+			+	+
Pakzad et al. [62]	2011	83	One region only				+						+
Nahm et al. [61]	2011	492	One region only	+						+	+
Husebye et al. [64]	2012	12	One region only	+
Vallier et al. [69]	2013	1005	Multiple regions	+	+	+	+			+	+		+
Vallier et al. [70]	2013	1443	Multiple regions	+	+	+	+					+
Stahel et al. [68]	2013	112	One region only				+						+
Dienstknecht et al. [67]	2013	165	One region only	+						+	+	+	+
Abrassart et al. [65]	2013	70	One region only		+							+	+
Böhme et al. [66]	2013	47	One region only		+							+
Cantu et al. [71]	2014	2323	One region only	+
Park et al. [72]	2014	166	One region only				+						+
Steinhausen et al. [73]	2014	379	One region only	+						+		+
Vallier et al. [76]	2015	335	Multiple regions	+	+	+	+					+
Konieczny et al. [74]	2015	38	One region only				+			+		+
Morshed et al. [75]	2015	2949	One region only	+							+
Reich et al. [77]	2016	376	Multiple regions	+	+	+	+					+
Glass et al. [79]	2017	294	Multiple regions	+	+	+	+	+	+				+
Byrne et al. [78]	2017	6948	One region only	+
Pape et al. [7]	2019	3668	Multiple regions	+	+	+	+	+	+	+		+	+
Devaney et al. [80]	2020	1270	Multiple regions		+	+						+
Tan et al. [85]	2021	103	Multiple regions	+	+			+		+		+
Denis-Aubrée et al. [81]	2021	201	One region only	+						+	+		+
Feldman et al. [82]	2021	96	One region only	+						+	+	+	+
Höch et al. [84]	2021	989	One region only		+							+
Flagstad et al. [83]	2021	328	One region only	+						+	+	+	+

Open in a separate window

+=Anatomic region/factor in focus

Open in a separate window

Fig. 2

Number of publications per anatomic region, focus on only one vs. multiple regions

Table 2

Number of publications per anatomic regions in respect to focus on one/multiple regions and time periods

	n=	Femur	Pelvis	Acetabulum	Spine	Tibia	UEX
Total	74	62 (83.7%)	22 (29.7%)	11 (14.9%)	15 (20.3%)	15 (20.3%)	10 (13.5%)
Focus on
One region only	50	40 (80%)	5 (10%)	0	5 (10%)	0	0
Multiple regions	24	22 (91.7%)	17 (70.8%)	11 (45.8%)	10 (41.7%)	15 (62.5%)	10 (41.7%)
Time period
1982–2009	44	42 (95.5%)	8 (18.2%)	3 (6.8%)	5 (11.4%)	11 (25%)	7 (15.9%)
1982–1989	8	8 (100%)	1 (12.5%)	1 (12.5%)	1 (12.5%)	2 (25%)	2 (25%)
1990–1999	20	19 (95%)	4 (20%)	2 (10%)	4 (20%)	6 (30%)	3 (15%)
2000–2009	16	15 (93.8%)	3 (18.8%)	0	0	3 (18.8%)	2 (12.5%)
2010–2021	30	20 (66.7%)	14 (46.7%)	8 (26.7%)	10 (33.3%)	4 (13.3%)	3 (10%)

Open in a separate window

Percentages are calculated within the corresponding strata (i.e., % of the total number “n”). Multiple mentions are possible

UEX upper extremity

Femur fractures were most frequently addressed and 62 (83.7%) of the 74 studies list it as a relevant fracture—40 studies focused exclusively on the femur. While some studies investigated the role of femoral fractures on the systemic inflammatory response as well as chest or brain injury, other studies compared the effect of immediate versus delayed surgical intervention or compared the effects of reamed and unreamed intramedullary nailing, plate osteosynthesis, and closed reduction external fixation.

Pelvic fractures were the second most common fracture type to be included in this synthesis. Twenty-two (29.7%) studies investigated pelvic fractures in the polytraumatized patient and five studies focused pelvic fractures exclusively. Overall, 14 studies were published after 2010. While some of these studies were concerned with the timing of fracture fixation, other studies focused on the role of instable pelvic ring fractures as the cause for hemodynamic instability.

Acetabular fractures were included in 11 (14.9%) studies, most of them in combination with injuries to the pelvic ring. Only three studies were published before 2010. All subsequent studies were focused on the role of early fracture fixation on the outcome in polytraumatized patients.

Spinal fractures were represented in 15 (20.3%) studies and 5 studies focused on spinal fractures exclusively. The majority of these studies (10 overall, 4 only focus) were published after 2010. Publications dealt with the timing of fracture stabilization in polytraumatized patients and its relevance for the outcome.

Tibia fractures were represented in 15 (20.3%) studies with no studies focusing on tibial fractures only. The vast majority of these studies (11/15) was published before 2010.

Upper extremity fractures were represented the least frequently and addressed in only 10 (13.5%) publications. There were no studies with an isolated focus on the upper extremity. The vast majority of studies (7/10) was published before 2010.

Determining factors

Besides the fractured anatomic region and general patient characteristics (age, co-morbidities), 67 out of 74 studies focused on at least one core factor, which determined the treatment strategy, or how patients were stratified. These determining factors could be organized into the four upper mentioned categories: Chest injury, TBI, Shock and “other”. An overview is presented in Tables ​Tables11 and ​and33.

Table 3

Number of publications per determining factor in respect to focus on one/multiple factors and time periods

	n=	Chest injury	TBI	Shock	Other
Total	74	42 (56.8%)	26 (35.1%)	25 (33.8%)	23 (31.1%)
Focus on
One factor only	39	18 (46.2%)	6 (15.4%)	7 (17.9%)	8 (20.5%)
Multiple factors	28	24 (85.7%)	20 (71.4%)	18 (64.3%)	15 (53.6%)
Time period
1982–1999	28	19 (67.9%)	10 (35.7%)	1 (3.6%)	4 (14.3%)
1982–1989	8	5 (62.5%)	2 (25%)	1 (12.5%)	2 (25%)
1990–1999	20	14 (70%)	8 (40%)	0	2 (10%)
2000–2021	36	23 (63.9%)	16 (44.4%)	24 (66.7%)	18 (50%)
2000–2009	16	10 (62.5%)	9 (56.3%)	8 (50%)	7 (43.8%)
2010–2021	30	13 (43.3%)	7 (23.3%)	16 (53.3%)	11 (36.7%)

Open in a separate window

Percentages are calculated within the corresponding strata (i.e., % of the total number “n”). Multiple mentions are possible

TBI traumatic brain injury, other spinal cord injury, abdominal injury or open fracture

Chest injury was mentioned as a determining factor in 42 publications and 18 publications focused on chest injury only. Of these 18 publications, 14 were published before 2000. Chest injury remained an important factor after 2000 (23 mentions), though seldom as an isolated entity.

TBI was mentioned in 26 publication and 6 publication focused on TBI as the only determining factor. While the number of mentions increased after 1990, there were no other apparent changes over the years.

Shock was mentioned as a determining factor in 25 publications and 7 publications focused on shock as an isolated entity. Interestingly, there was only one publication mentioning shock before 2000.

Other relevant injuries were included in 23 publications. These factors were mostly injury specific, meaning SCI in connection with spinal fractures and open fractures in connection with long bone fractures. Most publications (18) were published after 2000.

Changes over time

We were able to identify relevant changes over time in both the anatomic regions of interest and the determining factors. They are presented in Tables ​Tables22 and ​and33 and Figs. 3 and ​and4.4. In regards to determining factors, there has been a remarkable development since the change of the millennia. While TBI and chest injuries remained important determining factors, shock and other relevant injury-specific factors were mentioned a lot more commonly. Furthermore, many publications focused on multiple factors instead of only one. In regards to anatomic regions of interest, the changes occurred mostly since 2010. We noted a steep increase in publications on pelvic, acetabular and spinal fractures, while tibial and UEX fractures decreased. Yet, the femur still remains the most common anatomic region of interest.

Open in a separate window

Fig. 3

Number of publications per anatomic region sorted by time period

Open in a separate window

Fig. 4

Number of publications on determining factors sorted by time period

Discussion

In the polytrauma patient, fracture management continues to be of crucial importance, which is represented within the large trauma registries. The German Trauma Registry [TR-DGU], for example, showed that 28.2% of seriously injured patients between 2018 and 2021 suffered severe injury (AIS≥3) to extremities and pelvis, while injuries to the spine were present in 29.6% of cases [86]. The anatomic region of a given injury plays a pivotal role for the surgical treatment strategy. Still, within the last four decades it seems that the focus in the treatment of polytraumatized patients moved beyond considering the injured anatomic region. Instead, surgical treatment strategies are increasingly determined by the local and systemic reaction to trauma and resuscitative efforts as well as individual patient and injury characteristics. Our systematic review demonstrated the following findings:

1. Despite the presence of multiple injuries in the patients addressed in each publication, most of the early publications investigated only a single anatomic region, commonly the femur.

2. Over time, the focus moved away from femur fractures only and pelvic, spine and acetabular fractures are now routinely investigated.

3. In addition to the location of the fracture, in most studies surgical treatment strategy was determined by considering other relevant factors such as chest trauma, TBI or hemodynamic stability.

4. Regarding these determining factors, most authors focused on chest and brain injury before the turn of the century, while afterwards there is an increased focus on hemodynamic stability and other, injury-specific factors.

In terms of the first main finding, femur fractures have long been under special attention due to complication rates associated with their treatment. As delayed fixation and prolonged traction was known to be associated with an increased risk of fat embolism syndrome [18], many early publications dealt with issues of co-factors that may affect the timing of the femoral nailing, especially chest and head injuries [4, 31, 36]. Over time, fractures of the femur were included in all fixation concepts and stabilized in a definitive (intramedullary nailing) or at least temporary (external fixation) fashion. Femur fractures were at the focus of numerous investigations and prospective randomized trials on the merits of damage-control strategies in polytrauma patients [52, 87]. Despite these historical aspects, however, it remains undisputed, that a femur fracture is an important and challenging “major fracture”, and implies high potential for relevant blood loss, systemic inflammatory response and pulmonary complications.

Our second main finding is likely the explained by the improved understanding of the physiology and an increased focus on the role of fracture reduction as a means of surgical resuscitation [84]. This applies especially to the reduction of the pelvic ring to stop infra/retroperitoneal bleeding, as well as realignment of the spinal column and decompression of the spinal cord to treat spinal shock. In case of the pelvis, these might even be performed percutaneously through supraacetabular fixators or sacroiliac screws [8]. As the reduction of the pelvic ring can largely improve the patient’s hemodynamic stability, we feel the importance of pelvic instability may even be underrepresented in the literature. The high potential of pelvic fractures to influence the physiological response can drive surgical treatment strategies and would therefore determine them as “major fractures”.

Another important explanation may be a more patient centered approach and increased focus on favorable functional outcomes through early fixation of pelvic, spinal and acetabular fractures [69]. With the sustained advances in resuscitative and surgical strategies and improvements in training and infrastructure for first responders, many patients can be effectively stabilized despite severe injuries [7, 76, 79]. In such cases, early stabilization of pelvis, acetabulum and spine has shown improved outcomes, thanks to early mobilization and improved patient positioning [69]. This increased focus on the functional outcome has become even more important in case of spinal fractures with spinal cord injuries. Even in borderline patients, these fractures can often be stabilized at highest priority, to establish the best possible neurological/functional outcome [72].

In view of the third and fourth finding, the high importance of the physiological response for surgical decision-making becomes even more apparent. Therefore, it is also very plausible, that many recent publications prioritize injuries to trunk and femur over tibia and the upper extremities. This increased focus on the physiology is also reflected in the grading of an injury (i.e., in the 2015 AIS) [9]. Here, vascular injuries have long been classified independently and are usually scored higher than fractures due to the high potential for a physiological derangement. For fractures, the addition of indicators for potential severe soft tissue injury and/or blood loss might also be helpful. In other scoring systems, such as chest trauma scores, the pure description of osseous injuries, such was serial rib fractures or flail chest was outweighed by adding soft tissue injury (lung contusion) and physiological parameters (i.e., Horovitz ratio) [88]. In the comparison of polytrauma scores, the sole description of anatomic variables appeared to have the weakest predictive value [89]. The addition of physiologic variables led to the Berlin definition of polytrauma, which provides a prediction for patients with a relevant risk of mortality [2]. Further determination of scores demonstrated that the predictive value of an isolated physiologic parameter can be improved by adding further physiologic variables [90]. Therefore, it might also be worth to consider including parameters indicative of patient physiology in the definition of the “major fracture”.

Furthermore, the authors feel that the importance of open fractures, fractures with (neuro-)vascular injuries, fractures with severely contused soft tissues, and amputations have been underrepresented in the literature. It is common practice to address these injuries immediately and with high priority, given the patient’s physiology allows for a damage-control intervention [8].

Finally, one can discuss whether our method used to approximate the definition of “major fractures” in this article is adequate; and there are certainly valuable methods that should be considered in the future to expand on our findings. One important next step to consider would be large registry studies, which could evaluate how different fractures effect the patient’s physiological response.

We feel, however, that the scientific literature gives a good approximation of what is relevant in the clinical field at a given time, which is nicely demonstrated by the increased focus on hemodynamic stability and surgical resuscitation. Therefore, our approach represents an adequate start for further investigations.

Limitations

We are aware that our study has certain limitations. One of them is certainly the inaccessibility of certain references. This especially concerns some older references, which would have given an important insight into the rationale behind the treatment of polytrauma patients almost 50 years ago. A second important limitation was the large heterogeneity within the included studies in respect to study populations, study designs and outcome parameters. Therefore, categorization with respect to anatomic regions and determining factors was performed by the authors through qualitative interpretation, while an unbiased quantitative synthesis of the data was not possible. A third issue was the limited ability to account for or quantify soft tissue injury, which was partly due to the delayed perception of the importance of soft tissue injuries in the literature but also due to limited documentation.

Conclusion

Our understanding of the effects of fractures and their operative stabilization on the systemic response and the overall outcome in the polytrauma patients has advanced over the years. While the anatomic regions still plays an important role in determining a “major fracture”, other factors have been included to determine surgical priority. Physiological data, which were underrepresented until the 2000s are now readily available. A weighting between different fractures based on their effects on the physiological response or possible complications, however, has not yet been performed. Furthermore, there is only limited information on the role of soft tissue status or the degree of fracture dislocation or classification. These considerations might prove beneficial in finding a universal consensus. Currently, it appears that all relevant pelvic, spinal and lower extremity fractures should be considered “major fractures” and that the inclusion of the physiological response might be appropriate. The role of soft tissue damage and neurovascular injuries remains to be determined.

Abbreviations

Funding

Open access funding provided by University of Zurich.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

References