1 Introduction

The ankle-brachial index (ABI) is a highly specific and sensitive clinical screening metric that can be used to screen for vascular injuries in the setting of lower extremity trauma.^1–31VRFs, Vascular Risk Factors.2ATPFs, Acute Tibial Plateau Fractures.3VRF, Vascular Risk Factor. This screening tool is often used to evaluate penetrating lower extremity trauma when hard signs of limb ischemia are not overtly present. In such cases, the ABI can be used to determine if patients need further workup with advanced diagnostic imaging such as computed tomography angiography (CTA). The ABI is also frequently used when providers have concern for compartment syndrome or acute vascular injury. Patients presenting with acute tibial plateau fractures are known to frequently develop compartment syndrome and carry a high rate of vascular injury depending on fracture injury pattern.² In such patients, an ABI of less than 0.9 is often used by orthopaedic providers to guide further diagnostic strategies and determine if vascular insult is truly present.^1–3 This examination and cut-off value has been validated by Lynch et al. and again by Mills et al. and continues to be a benchmark for evaluation of lower extremity trauma and vascular injury.^2,3 However, it remains unclear whether comorbid conditions that affect perfusion and peripheral vasculature confound the detection of subtle vascular injury, as patients with comorbid vascular pathologies already have an ‘injured’ vascular system.

Within trauma and cardiovascular literature, an ABI value of 0.9 is well-defined as the lower limit of normal for healthy adults.^1,4–7 Below a value of 0.9, patients are not only far more likely to have vascular injury from acute trauma, but also more likely to have significant peripheral vascular disease (PVD) and develop secondary pathologic chronic limb perfusion. Mills et al. reported that among the trauma patient population, the average ABI for patients presenting without vascular injury ranged from 0.9 to 1.2.³ However, their study focused on the lower limit of normal ABI values, while the upper limit of normal was not as clearly defined. Ishida et al. divided ABI values in the normal-to-high range into three subpopulations: “normal” (0.9–1.19), “high-normal” (1.2–1.39), and “high” (>1.4) and found that patients with “high” range ABI values were associated with the greatest risk of mortality.⁴ These findings are in agreement with epidemiological reports on patients within the general population, wherein “high” ABI values (>1.4) were found to have increased rates of morbidity and mortality secondary to PVD.^5,6 While such studies provide a wealth of knowledge for the evaluation of chronic limb ischemia in the general patient population, there remains a scarcity of data to guide providers on “normal” ABI values within the trauma patient population. Still less is known regarding the “upper” limit of normal within trauma patients, therefore the evaluation of trauma patients with vascular risk factors in the acute setting remains ambiguous at best.

By our own institutional protocol, most low energy tibia plateau fractures, all high energy tibia plateau fractures, lower extremity gunshot wounds, knee dislocations, and ‘floating knees’ (fractures above and below the knee) receive an ABI on initial evaluation. We classify high energy injuries as falls greater than 8 feet, motor vehicle accidents, motorcycle accidents, crush injuries, pedestrians struck by a vehicle, young (<40 years old) sports injuries and ballistics injuries. At our Level 1 Trauma Center, if an ABI value is less than 0.9 in this patient population, we work them up with a CTA study or angiogram. However, the idea of falsely elevated ABI values in certain subgroups of patients led the authors to question if a threshold value of 0.9 can be ubiquitously applied to all trauma patients, or whether there needs to be a different cutoff value for this patient subgroup. Patients with diabetes mellitus (DM), chronic kidney disease (CKD) and PVD who have sustained chronic vascular damage are proposed to be most likely to yield skewed readings for ABI.^8,9 This mechanism has been explored and is believed to be secondary to the development of less compressible arteries that in turn, falsely elevate ABI values.^8,9 However, although commonly discussed amongst trauma providers, our review of the literature revealed little to no evidence regarding the impact of comorbid vascular pathologies such as DM, CKD, and PVD and the use of ABI to evaluate for subtle vascular injuries in acutely traumatic patients. The authors were left to wonder if this patient subgroup may be at higher risk for missed vascular injuries due to falsely elevated ABI, and as such, if these patients may benefit from a CTA at a different ABI threshold compared to the general patient population.

The goal of our study was to perform a retrospective review of patients who sustained acute tibial plateau fractures and determine if there was a significant difference between the ABI of patients with and without vascular risk factors (DM, CKD, and PVD). We further aimed to determine if other factors such as age, open versus closed injuries pattern, mechanism, energy level, Schatzker type, or concomitant ipsilateral limb injury influenced the presenting ABI. To our knowledge, this study is the first to formally examine the effect of comorbid vascular pathology on ABI values in the setting of acute tibial plateau fractures.

2 Methods

3 Results

490 tibial plateau injuries were identified using our hospital EMR data collection since 2014. Of those, 282 of those injuries had specific ABI values charted at the incident time of arrival. 46 patients were identified as having the risk factors DM, CKD, or PVD within this cohort (Fig. 1). The average ABI on arrival of the risk factor (RF) group was 0.95 ± 0.15 versus non-risk factor (NRF) group 1.0 ± 0.15 and there was no statistical significance between the mean ABI on arrival though it did near significance (p = 0.06) (Table 1). Furthermore, there was no difference between the number of RF patients with a reported ABI less than 0.9 (28.3%) versus NRF patients (14.1%) with p-value of 0.65. We further examined if the patient groups were more likely to present with an ABI <0.9 or >1.1. 0.9 is a universally accepted cutoff before a CTA is generally recommended, whereas an upper limit of 1.1 is not as clearly defined. Most of the literature comments on either 1.1 or 1.2 as the upper limit of normal,^3,5,6 but to increase sensitivity we chose 1.1. This subgroup analysis did not reveal any significant difference between the two patient populations when we looked at ABI <0.9 (p = 0.65) or >1.1 (p = 0.45) (Table 2). Demographic analysis revealed there was a statistically significant difference in average age between the two groups, 57.3 ± 11.0 years in the RF group versus 47.8 ± 14.5 years in the NRF group (p-value < 0.001). There was no statistical significance in sex; 48% female in the RF group versus 40% in the NRF group (p = 0.373) (Table 3). There was also no statistical significance to whether the injury was open versus closed, 13% open in the RF group versus 12% open in the NRF group (p = 0.849). When reviewing the patients who had an ABI charted as “>1”, none had a vascular injury and only one of seventeen required a four-compartment fasciotomy and that patient was in the NRF group. Of the patients who required vascular intervention or four-compartment fasciotomy (n = 58), none had the risk factors of interest.

Close

Fig. 1 Patient selection flow chart.

Export to PPT

Within each subgroup, there was no statistical significance of mean ABI between low and high energy mechanism presentations. In the NRF group, the average ABI was 1.0 ± 0.15 in the high energy versus 0.99 ± 0.15 in the low energy group (p = 0.96). In the RF group, the average ABI for high energy mechanisms was 0.99 ± 0.11 versus the low energy average of 0.89 ± 0.17 (p = 0.06)). There was a statistical significance between low energy mechanisms in RF patients (mean ABI 0.89) and NRF patients (mean ABI 0.99) with p-value of 0.02 (Table 1).

Having an ipsilateral limb fracture (hip fracture, femoral shaft fracture, distal femur fracture, tibial shaft fracture, or ankle fracture) did not prove to have a statistically significant mean ABI difference compared to RF and NRF groups. More specifically, patients with a “floating knee” injury did not have a statistically significant ABI on arrival when compared to patients without an ipsilateral injury (p = 0.89). When comparing different Schatzker injury patterns using analysis of variance (ANOVA) there was a statistically significant difference of mean ABI on arrival (p = 0.048). Games-Howell post-hoc analysis of mean ABI on arrival indicated the difference between type 5 (bicondylar split) and type 6 (metaphyseal-diaphyseal separation) was statistically insignificant (p-value = 0.03) (Table 1).

4 Discussion

Patients with underlying vascular damage are known to raise concern for falsely elevated ABI in the evaluation of acute arterial injury secondary to lower extremity trauma. This potential confounder is often discussed amongst the providers of patients with underlying vascular disease, yet little data exists to support this concern within the literature. In our analysis, of patients with acute tibial plateau fractures, there was no difference in ABI values between patients with or without vascular risk factors. This was irrespective of patient demographic, as there was no difference in our patient populations regarding those who presented with a low (<0.9) or high ABI (>1.1). Such data would suggest that trauma patients with vascular risk factors are no more likely to have a falsely elevated, or low, ABI than those patients without vascular risk factors.

There are several limitations to the present study, such as its retrospective nature. As in most retrospective studies, it is difficult to know the exact details of how the history and exam was performed but we attempted to obtain a significant number of patients who were evaluated by our orthopaedic surgery team in attempt to standardize patient evaluation. However, measurement of the ABI is highly operator-dependent and there remains a lack of consensus regarding measurement technique within the literature.^1,3 It is therefore possible that various factors such as type of equipment used and operator-experience may also have introduced some degree of inter- and intra-observer error. Furthermore, a significant amount of the original 490 patients with tibial plateau fractures did not have a specific ABI value recorded in the medical record, and therefore far fewer patients were included in our analysis. Such selection bias limits study leading to the possibility that our data may approach clinical significance with a larger cohort. We did not stratify the ABI results based on severity of disease, for example chronicity of diabetes, hemoglobin A1c values, severity of CKD and PVD which may illuminate further differences in ABI values and possibly identify patients with more severe vascular disease where falsely elevated ABI values may hold true.

After examining the mean presenting ABI for each group, we then proceeded to subgroup analysis to try and determine other specific factors such as energy mechanisms and injury pattern that may result in a statistically different ABI. High and low energy mechanisms could reveal possible ABI differences owing to the hypothetical differences in soft tissue injury with higher kinetic energy and therefore more likely to have vascular compromise. We generally classify high energy at our institution as falls greater than 8 feet, motor vehicle crashes, motorcycle crashes, sports related injuries in patients under 40, pedestrian struck injuries and ballistic injuries. Low energy mechanisms were almost always ground level mechanical falls in elderly or low energy twisting injuries. Most patients in the NRF group (77%) had high energy mechanisms but there was no statistical significance between mean ABI in high versus low energy mechanisms. In the RF group the majority were also high energy (63%) but again there was no statistical difference in mean ABI though it approached statistical significance (p = 0.059). The only significant difference existed when comparing the RF to NRF groups regarding low energy mechanisms, 0.89 versus 0.99, respectively with p = 0.019. There was no statistical significance when comparing high or low energy mechanisms within each RF and NRF group. Therefore, based on our data there does not appear to be an associated between mechanistic energy level and need to pursue diagnostic vascular studies (i.e., CT angiogram) based on presenting ABI unless the value is < 0.9 as is standard.

The Schatzker classification exists to provide structural description of tibial plateau fractures and help identify subtypes that may be length unstable and possibly at risk for neurovascular compromise.¹⁰ Most commonly type 4 (medial column), type 5 (bicondylar split), and type 6 (metaphyseal-diaphyseal dissociation are associated with “length unstable” injury patterns wherein the knee has equivalently dislocated and places the neurovascular structures at risk. These subtypes are also more likely to develop compartment syndrome secondary to the significant displacement and commonly higher energy level. We attempted to examine whether Schatzker type was associated with a statistically significant mean ABI on arrival. Through post-hoc analysis after ANOVA, the only significant difference existed between type 5 (mean ABI 1.09) and 6 (mean ABI 0.97). None of the other injury patterns were significantly different. This analysis is of course limited by accurately identifying the injury pattern base on radiographs and advanced imaging and attempting to “fit” the injury pattern into a specific subtype. There were of course multiple (n = 12) injury patterns that did not align well with the Schatzker classification but based on the classification system used were a best fit. Our breakdown of Schatzker type from 282 patients was as follows: 3% type 1, 24% type 2, <1% type 3, 12% type 4, 2% type 5, and 58% type 6.

Another patient variable suspected to possibly influence the risk for a vascular injury and thus significantly different mean ABI was ipsilateral extremity injuries, most specifically floating knee injuries. Floating knee injuries as well as floating elbow injuries are known to be risk factors for compartment syndrome in extremities, therefore we suspected this could be an identifiable variable possibly reflected in the presenting ABI.^11,12 Ipsilateral injuries were considered as involving any part of the femur (femoral head to distal femur) or any part of the tibia distal to the articular surface (tibial shaft to malleolar ankle fractures). On analysis there was no difference in presenting mean ABI of patients with an ipsilateral extremity injury versus the RF or NRF patients who did not have an ipsilateral injury. If we isolated the ipsilateral injuries to just floating knees (i.e., femoral shaft fractures) there still was no statistically significant difference. Therefore, despite the concern for compartment syndrome in patients with ipsilateral injuries this was not statistically reflected in the mean ABI of these specific populations.

One of the more interesting aspects found was of all patients who required a four-compartment fasciotomy or vascular intervention, none of them had the risk factors in question. As the data showed, the risk factor population was statistically older and despite most of the RF group being high energy mechanisms, they overall had a higher percentage of low energy mechanisms. This could reflect the patient population sampled but could represent a possible underlying pathophysiologic change in older patients with vascular disease. These factors combined could result in an increased fascia compliance allowing for more swelling within the compartment as well has sarcopenia leading to a less robust inflammatory and swelling response. This is hypothetical but could explain the possible difference between the RF and NRF groups and why no RF patient required a fasciotomy or vascular repair. Further biological studies comparing fascial compliance and muscle swelling potential could elucidate this difference.

5 Conclusions

There was no statistical significance in the presenting ABI value of patients with DM, CKD, or PVD versus those without these risk factors sustaining an acute tibial plateau fracture. Further, there was no difference in the proportion of patients presenting with an ABI less than 0.9 or greater than 1.1, sex, or open versus closed injury. Interestingly, none of the patients requiring a four-compartment fasciotomy or vascular intervention had the risk factors in question. Only with subgroup analysis could statistically significant mean ABI values be found between low energy mechanism presentations of RF and NRF patients and those with Schatzker type 5 and 6 injury patterns. This study suggests patients with vascular risk factors have a similar ABI as those without vascular risk factors.