Introduction

Venous thromboembolic events (VTE) (including deep vein thrombosis [DVT] and pulmonary embolism [PE]) are known to further accentuate the complexity of managing patients with pelvic and acetabular fractures (PAFs). VTEs are associated with significant morbidity and mortality of the affected population. Patients who were diagnosed with DVT experienced significant reduction of their health-related quality of life (HrQoL) in addition to the long-term higher mortality rates [1]. The overall 30-day mortality rate of PE can reach 11.4% [2].

PAFs are high-energy complex injuries, usually associated with other fractures and require multidisciplinary team approach in highly specialised trauma centres. In spite of this level of care, exceptionally high rates of thromboembolism are still encountered by the treating surgeons. Incidences of up to 41% percent were reported for DVT and 21.7 percent for PE [3, 4].

Aiming to guide the process of VTE Prophylaxis in those patients, we present this review with three main goals: (A) highlight the incidence of VTE in PAFs and identify the risk factors, (B) investigate the screening methods available in the current literature to assess their sensitivity and validity and examine their impact on the diagnosis and treatment, (C) search the VTE prophylaxis protocols to extract best practice guidelines for PAF patients.

Materials and methods

Search strategy

We performed a systematic review following the preferred reporting items for systematic reviews and meta-analysis (PRISMA) guidelines [5]. The online published databases of PubMed and EMBASE together with the Cochrane library were searched from inception to May 2021. PubMed and EMBASE databases were individually searched through the OVID Online research platform, while the Cochrane library was searched through its online website. Our search included EPUB ahead of print and in-process and other non-indexed citations. The MeSH terms searched were “Pelvis/Pelvic”, “Acetabulum/Acetabular” AND “Fracture”, “Thrombosis”, “Embolism”, “Thromboembolism”, “Pulmonary Embolism” and “Filter”. Subsequently, the terms were paired together to create lists of results. All search terms were “Exploded”; then all subheadings were included. Reference lists of the included studies were also screened for relevant publications.

Eligibility criteria

All publications pertinent to thromboembolic events in relation to PAFs were reviewed. Level V publications were ruled out. Search was limited to English literature and human studies.

Study selection and critical appraisal

The titles and abstracts of the list of studies were independently reviewed by each of the authors to confirm eligibility. Subsequently, final agreement on included publications was reached by discussion among all authors. Selected studies were then assessed for the risk of bias using MINORS criteria for non-randomized trials and RoB-2 tool for randomized trials [6, 7].

Study classification

Selected studies were subsequently classified into studies pertaining incidence of VTEs, screening for thrombosis, and methods of thromboprophylaxis including the use of inferior vena cava filters (IVCF). This created 4 groups of listings to aid data extraction and review narration.

Data extraction

All manuscripts, but one [8], were downloaded and data extracted including number of patients, incidence of DVT, incidence of PE, study methodology, and conclusion.

Data from different studies were downloaded (or requested) in an attempt to establish fracture types related to VTEs.

For screening studies, methods of screening were considered, and for prophylaxis studies, the prophylaxis methods were listed.

Results

Numbers of records retrieved per database together with the search strategy are presented in Fig. 1. In addition to the previously excluded records, one study had to be ruled out because of methodology issue failing to separate and present pelvic trauma patients from other trauma patients.

Fig. 1
figure 1

PRISMA search flow chart

Full size image

In total, 28 studies were found relevant to the reviewed topic. These included one prospective randomized study [9], four comparative studies [10,11,12,13], and four registry data-based studies [14,15,16], and the rest were cohort studies. All studies were assessed for the risk of Bias. RoB-2 can be used to assess two studies, both of which were found to have a high risk of bias.

The rest of studies were scored according to MINORS criteria, and the most common score noticed was 7 (range [5,6,7,8,9,10]:).

For the purpose of the review, studies were grouped into four categories. Further critique and details of the studies involved will be presented in each group.

Incidence of thromboembolic events in pelvic and acetabular fractures

Ten studies, published between 1993 and 2020, reported the incidence of VTEs. Eight were cohort studies, four prospective and four retrospective, and two were registry/database retrieval reports. Collectively, all studies included 89,046 patients with PAFs. Minimum follow-up was 90 days [14] and varied between studies up to a maximum of seven years [17]. Follow-up was not documented in two studies [3, 18]. The reported incidence of DVT was highly variable starting from as low as 0.21% [14] and reaching up to 41% [17]. Other reported values are listed in Table 1. Similarly, incidence of PE ranged from zero [21] and up to 21.7% [22].

Table 1 Incidence of VTE in pelvic and acetabular fracture patients
Full size table

Multiple factors could have contributed to the noted significant variation. Firstly, the diversity of diagnostic tools used by the investigators could result in either over- or under reporting of the VTE events. Secondly, different protocols were used to identify patients for investigation; while some studies screened all patients [3, 23], others were selective in choosing patients for investigation [18], and in a single study, by Niikura et al., two different protocols were applied to their patient cohort [17]. Thirdly, our search produced three national database publications [14,15, – 16]. These were noted to report significantly lower rates of VTEs.

Screening for VTE in pelvic and acetabular trauma patients

Nine studies, published between 1990 and 2020, including 1360 patients, presented different screening modalities, all of which aiming at early detection of VTE in PAFs (Table 2). These included duplex ultrasound scan (USS), contrast venography, venographic contrast computed tomography (CTV), and magnetic resonance venography (MRV) [10, 21, 23,24,25,26,27,28]. All studies, but one [10], are prospective cohort studies. Follow-up intervals varied from two weeks post-operative [24] and up to six months [23].

Table 2 Screening for VTE in pelvic and acetabular trauma patients
Full size table

Some investigators used ultrasound scanning protocols to guide future management [10, 24,25,26]. These included either further investigations [10, 26] or initiation of the anticoagulation protocols [23,24,25]. The later was followed in three studies, in which a total of 638 patients were screened pre- and post-operatively [23,24,25]. Patients with pre-operative DVT underwent IVC filter insertion, while therapeutic anticoagulation was prescribed for those with post-operative proximal DVT. None of the studies presented their rates of DVT, PE, or VTE related mortality prior to the screening program. Elnahal et al. [24] reported that their protocol has changed the management in five patients; therefore, none of them developed PE. Nevertheless, it was noted that in all three studies, USS screening programs could not prevent post-operative PE which was diagnosed in ten patients (1.6%). In this scenario, PE could be either due to progression of DVT that was not detected on USS or a de novo PE (DNPE) triggered by the acute inflammatory response [29].

White et al. used the USS for serial screening, and subsequent contrast venography was used to confirm the USS findings [26]. Borer et al. presented the difference between the incidence of PE in a group of screened patients (using USS and MRV) and another group of unscreened patients. At the end of the study, the incidence of PE was even higher in the screened group (2 versus 1.4%). All patients diagnosed with PE initially tested negative upon screening [10].

The literature available pertaining MRV as a screening tool was controversial. Two studies, same authors, reported it to be sufficiently sensitive, noninvasive test and superior to invasive venography. They reported one incidence of PE among 146 patients (0.6%) [28, 30]. On the other hand, Stover et al. used MRV and CTV for screening and identified unacceptably high false positive results of both tests, 100% and 50% respectively, in comparison to invasive pelvic venography [27].

VTE prophylaxis in pelvic and acetabular fractures

Ten papers published between the years 1994 and 2021, including 2836 patients, have specifically investigated VTE prophylaxis in PAFs. Two studies were retrospective [21, 31], one database analysis [20], and the remaining was prospective [9, 11,12,13, 22, 23, 32]. All studies followed up patients only during the acute episode of admission and/or shortly after, with the longest follow-up being 6 months [23]. All studies, apart from one [33], included adult patients (Table 3).

Table 3 Methods of VTE prophylaxis in pelvic and acetabular trauma patients
Full size table

Patient age

Greenwald et al. analyzed the incidence of VTE in pediatric patients with PAFs [33]. Of their large cohort, 8.8% (948 patients) received some form of prophylaxis and the incidence of DVT was only 0.17% with no detected PE events. They demonstrated the low risk of VTE related morbidity and mortality in this age group.

Mechanical prophylaxis

Two studies were primarily focused on mechanical VTE prophylaxis. Stannard et al. [7] performed a randomized controlled trial comparing the high-pressure pulsatile mechanical to the standard low-pressure sequential calf compression devices. A trend towards a lower incidence of DVT (9 versus 19%) was noted with pulsatile compression; however, the difference did not reach statistical significance.

Another cohort study reported, unsurprisingly, high rates of DVT and PE (18 and 12% respectively) in PAF patients following isolated use of mechanical prophylaxis (graduated compression stockings and intermittent pneumatic compression devices) [22].

Timing of pharmacological prophylaxis

Two studies questioned the late administration of pharmacological prophylaxis [low molecular weight (LMWH) and unfractionated heparin (UH)] as a risk factor for VTE. Both have confirmed that administration after 48 h from arrival to the emergency department was associated with significantly increased risk of VTE in PAF patients [13, 31]. Using LMWH for prophylaxis, Steele et al. [35] have shown that administration within 24 hours from the injury, or achieving haemodynamic stability, was associated with significant reduction of the incidence of DVT.

With the increasing use of the direct oral anticoagulant (DOAC), rivaroxaban, Monzon et al. [12] visited the same question. They found that administration of rivaroxaban within 24 h was associated with significant reduction of the incidence of DVT without increased risk of intra- or post-operative bleeding.

LMWH versus DOACs

Hamidi et al. [20] were the only group to compare the use of LMWH to DOACs in PAF patients. They presented database analysis of a large number of non-operatively treated patients. They identified that DOACS were associated with significant reduction of the risk of DVT, but not PE, without increasing the risk of bleeding or mortality.

Dosage monitoring

Constantini et al. [32] demonstrated the advantage of using plasma anti-Xa in trauma patients to monitor the therapeutic effects of LMWH. Only 29.5% of their patients were found to have the expected circulating therapeutic levels of anti-Xa. Hence, they recommended using higher prophylactic doses of LMWH (more than 30mg twice daily). Alternatively, anti-Xa can be routinely assessed for dose adjustment in trauma patients.

Inferior vena cava filter (IVCF) for thromboprophylaxis in pelvic and acetabular fractures

Our search revealed three studies, published between 1992 and 2019 and included 370 patients. Two comparative studies [8, 36] examined two groups of patients, IVCF and non-IVCF; both were single centre studies with non-randomized allocation (Table 4).

Table 4 Use of IVCF in pelvic and acetabulum fractures
Full size table

The first study (Cohen-Levy et al. [36]) failed to show significant reduction of the incidence of PE. They have also noted an associated nonsignificant rise in the incidence of DVT among IVCF patients. Therefore, it could not confirm the benefit of the IVCFs.

In the second study, Webb et al. [8] showed a reduction of the incidence of PE among the IVCF group (0% versus 7%). However, 17% of the IVCF patients developed leg oedema. The oedema was severe enough to result in peripheral lower extremity tissue loss in one patient.

The third study, by Toro et al. [37], presented a four-year follow-up result of 88 patients who were preoperatively diagnosed with DVT requiring the insertion of IVCF. None of their patients developed PE or recurrent DVT. 7% developed bilateral lower limb swellings, and 1% suffered from post thrombotic syndrome affecting both lower limbs.

Discussion

Despite all advances in recent anticoagulation therapy and thromboprophylaxis, PAF patients still suffer from significantly high rates of thromboembolic events. The current literature highlights the fact that, despite its low rates, PE is challenging to predict and eliminate. Mortality rate due to untreated PE can reach as high as 30%, while in diagnosed and treated PE, it is 8%. Two of three patients with PE die within two hours after presentation [19].

The incidence of VTEs was highly variable in the current literature, and only one study reported zero PE in their patients; otherwise, both DVT and PE were unavoidable.

Risk factors for VTEs in PAF patients include lower limb external fixation and those with high Charlson comorbidity index [18]. Patients with high body mass index (BMI > 30) and those who underwent pelvic angioembolization were found to be at higher risk (up to 2.6 times) of VTE [13, 38].

When it comes specifically to PE, risks included obese (BMI > 40) males, with history of warfarin use, intensive care unit admission, high ISS scores (>15), and associated fractures above the knee [15, 18, 34].

Wang et al. reported that the incidence of DVT was higher among patients older than 60 years old, with associated injuries and who underwent late surgical fixation (> 2 weeks). In addition, they found that DVTs were significantly increased in patients with acetabular fractures more than those with pelvic ring injuries and that proximal DVT was significantly higher with complex than simple acetabular fractures [3].

Dwyer et al. highlighted the importance of close follow-up of those patients and immediate attention to any suspicious symptoms of VTE. In their cohort, 28% of DVTs and 23% of PEs happened more than 35 days after discharge [14].

Screening for VTE still cannot be coined as the best practice in those patients for many reasons.

Firstly, none of the screening methods has shown itself to be the gold standard technique. In spite of being noninvasive, economic, and easy to organize, USS was not sensitive enough to reduce the incidence of PE [10, 25]. Some authors used it as an initial screening method and used more advanced tests to confirm the finding [26]. In addition, PAF patients are commonly polytraumatized. Hence, an associated lower limb injury, cast, or external fixation can hinder the use of USS.

MRV is the most recent imaging modality for detection of VTEs. It has the advantage of being noninvasive and offering a detailed high-resolution picture of the venous tree (Fig. 2) [39].

Fig. 2
figure 2

Example of MRV Image demonstrating DVT of the left common iliac vein (left image) and the right common iliac vein (right image). [Obtained from Tamura K et al. MR Venography for the Assessment of Deep Vein Thrombosis in Lower Extremities with Varicose Veins. Ann Vasc Dis. 2014;7(4):399-403]

Full size image

Nevertheless, although it was considered very precise by some [30], it was found to be too sensitive and resulting in 100% false positive results by others [27]. Ironically, both studies based their conclusions by comparing MRV to invasive contrast venography. MRV also have the disadvantages of being an expensive study, not readily available, requiring patient transfer to another department, and it can be affected by implanted metal work.

Secondly, the screening process itself can be inconvenient. Those patients, frequently, have multiple visits to the operating theatres prior to the pelvic fixation; some of them would even attend the angiography suite for embolization prior to screening. Hence, the complexity of identifying the best time to perform the screening test.

In addition, a recent survey of the PAF surgeons of the Orthopaedic Trauma Association (OTA) showed that only 8.7% of the surgeons would obtain routine VTE screening for their patients on admission [40].

Both mechanical and pharmacological VTE prophylaxis were visited in multiple studies. Our review highlighted four key messages:

  1. (1)

    Both mechanical and pharmacological VTE prophylaxis are essential in PAF patients. Studies attempting the isolated use of mechanical methods have shown unacceptably high rates of VTE [4, 22].

  2. (2)

    Unless contraindicated, pharmacological agents must be started as soon as possible once haemodynamic stability is achieved, ideally within 24 hours of the patients’ attendance to the emergency department [11,12,13, 31].

  3. (3)

    Despite the preferred use of LMWH among orthopedic surgeons [40], two studies have confirmed the safety and efficacy of using DOACs in VTE prophylaxis for PAF patients [12, 20].

  4. (4)

    Caution needs to be exercised before the elaborate use of IVCFs for patients with PAFs. Two studies confirmed its benefit [8, 37] and one declined it [36].

Our review provided a broad presentation of the available literature on VTE in patients with PAFs. Limitations included level of evidence of the studies included (level II and below) and the absence of meta-analysis due to the heterogeneity of the studies.

High-level studies are still required to provide the best practice guidelines on VTE prophylaxis in patients with pelvic and acetabular fractures.