History Knee dislocations are limb-threatening injuries defined by loss of reduction in tibiofemoral articulation associated with disruption of at least one cruciate ligament. Although many clinicians associate knee dislocations with high-energy, traumatic injuries, it is important to note that about 40% of knee dislocations occur through low-energy mechanisms; this is more common in women and people with obesity [2, 5, 6]. The description and classification of knee dislocations have historically involved three variables: the energy level of the injury, positional displacement of the tibia, and damaged anatomic structures. The energy classification of a knee dislocation is a useful indicator of the severity of dislocation and probability of arterial and nerve damage [27]. However, although the energy level is useful as an initial evaluation tool for knee dislocations, it is inadequate as a means to classify these injuries. Positional displacement of the tibia with respect to the femur was first standardized in the locational classification of knee dislocations system developed in 1963 by Kennedy [9]. The Kennedy positional classification includes five categories: medial, lateral, anterior, posterior, and rotatory. Rotatory injuries can be further divided into anteromedial, anterolateral, posteromedial, and posterolateral injuries [9]. Despite the ease of use of this classification system, it is limited in its identification of the severity and extent of injury and cannot be used in the knee that has spontaneously reduced, as commonly occurs in patients with knee dislocations. To account for limitations of other classification systems, Schenck [22] developed his anatomic classification system for knee dislocations in 1994, based on ligamentous damage rather than the direction of displacement. More recently, Boisgard et al. [1] developed an anatomic classification system for knee injuries that includes bicruciate lesions without dislocation. However, Schenck’s system, including the addition of a fracture-dislocation category by Wascher et al. [28], has become the prevailing anatomic classification method for knee dislocations and is used routinely in clinical practice. Purpose Schenck developed his classification system to address limitations of previous classification systems and to specify the structures damaged. An ideal classification system is reproducible, facilitates clear documentation and communication, encompasses all pathologic types of deformity, guides treatment, correlates with prognosis, and is applicable to research. As knee dislocations include a wide range of injured structures, the severity of injury to the structure may vary widely. Also, many knee dislocations spontaneously reduce, leading to a high rate of misdiagnosis and improper initial treatment [16]. Therefore, a classification scheme using damaged anatomic structures rather than simply the direction of tibial displacement was intended to allow improved reproducibility, particularly in the case of spontaneously reduced dislocations. Classifying dislocations based on injured structures also allows for more detailed communication between clinicians. This is in contrast to the positional classification, which only provides information on the radiographic or gross appearance of the dislocated knee. Additionally, the anatomic information provided by the system was intended to guide treatment and assist clinicians preparing for ligament reconstruction or repair [22]. Finally, the classification system was designed to better predict neurovascular complications associated with knee dislocations. Injury to the popliteal artery is the most common vascular injury associated with knee dislocation, with an estimated prevalence of 18% in all knee dislocation, with studies reporting up to a 43% prevalence after high-energy dislocations [11, 17]. Arterial injuries are limb-threatening, with an estimated resultant amputation rate of 12% [17]. Neurologically, peroneal nerve damage associated with knee dislocation may lead to foot drop or loss of sensation in the foot, which is estimated to occur in 10% to 40% of knee dislocations [19]. Therefore, diagnosing and treating these serious complications is an important component of knee dislocation treatment, which Schenck intended to improve through his classification system. Description The Schenck classification of knee dislocation uses damage to the collateral and cruciate ligaments to distinguish knee dislocations based on severity, with higher grades in the classification involving a greater number of ligaments (Table 1). Knee dislocation I (KD I) is the least severe form of knee dislocation and is defined as a dislocation with a tear of only one cruciate ligament (ACL or PCL) and radiographic evidence of tibiofemoral dislocation. KD II dislocations are uncommon [4] and include tears to both cruciate ligaments (ACL and PCL), with functionally intact collateral ligaments. KD III is characterized by tears of both cruciate ligaments and one collateral ligament and is further subcategorized by which collateral ligament is involved. KD IIIM includes tears of the ACL, PCL, and medial collateral ligament, while KD IIIL includes tears of the ACL, PCL, and lateral collateral ligament. KD IV is characterized by tears of both cruciate ligaments and both collateral ligaments and is generally seen in high-energy injuries [24]. KD V is equivalent to a fracture-dislocation, which may be classified further using other systems, such as the Moore [18] or Schatzker et al. [21] classifications of tibial plateau fractures. Knee dislocations may then be subcategorized further to indicate associated artery or peripheral nerve injury. Popliteal artery injury is indicated with “C” and peripheral nerve injury, typically to the tibial or peroneal nerve, indicated with “N” [24]. Through this classification system, Schenck hoped to stratify dislocations based on the risk of arterial and nerve damage and to classify dislocations based on recommended treatment and prognosis [23].Table 1.: Description of Schenck classification criteria for knee dislocationsValidation The intraobserver and interobserver reliability of the Schenck classification is limited because it has not been studied directly. The small sample of patients with knee dislocations makes evaluating the reliability challenging. Therefore, the lack of validation studies of the classification system is a major limitation. Although reliability has not been studied directly for the Schenck classification, tests for isolated ligament integrity have been validated. These validation studies do not apply directly to knee dislocations. However, multiple ligamentous injuries are often involved in knee dislocations, making these tests important components of the evaluation process. The Lachman and pivot-shift tests, posterior drawer and posterior sag tests, and valgus and varus stress tests are used to assess the integrity of the ACL, PCL, medial collateral ligament (MCL), and lateral collateral ligament (LCL), respectively [7]. These clinical tests are then typically followed by MRI to further characterize the injury. Studies for the interobserver and intraobserver reliability of physical examination tests for isolated ACL disruption have varied, with the Lachman’s test as the most reliable with an intraobserver reliability of k = 0.29 to 1.00 and an interobserver reliability of k = 0.23 to 0.92 [13]. The posterior drawer test of PCL function has been found to be reliable, with an estimated accuracy of 96% [20]. MRI accuracy of diagnosing isolated ACL, PCL, MCL, and LCL tears has also been evaluated. The accuracy of MRI diagnosis of ACL, PCL, and LCL injury compared with arthroscopy was found to be 95%, 99%, and 95%, respectively [8, 10, 14, 15]. Evaluation of the accuracy of MRI diagnosis of MCL injury is less established, as it is generally a nonoperative injury. Overall, evaluation of an isolated single-ligament injury has been studied and shown to be reliable, particularly for ACL, PCL, and LCL injuries. Although these studies have not been evaluated in the context of a dislocated knee, it seems likely that these tests remain reliable in multiple-ligament injuries, such as in knee dislocations. The Schenck classification system has been evaluated to determine its utility to predict neurovascular complications, as well as its usefulness as a guide to postsurgical prognosis. A systematic review by Medina et al. [17] found the highest rates of vascular injury in patients with KD IV injuries, while Stannard et al. [26] found the highest rates of vascular injury in KD IIIL injuries. Neither study found evidence of popliteal artery damage in KD I or KD II injuries (Table 2). Furthermore, a study of 91 patients with knee dislocations found a strong association between KD IIIL injuries and peroneal nerve injury. Although only 36% of the 91 patients had KD IIIL injuries [19], the KD IIIL group comprised 69% of the patients with peroneal nerve palsy. Further, there is evidence that postoperative prognosis varies by Schenck classification. A study by Frosch et al. [4] found worse clinical outcomes among patients with KD IV injuries than among those with KD II and KD III injuries based on IKDC and Lysholm scores. They did not, however, find a difference in these outcome measures between KD II and KD III injuries. Similarly, Everhart et al. [3] found lower rates of return to work after KD IV and KD V injuries compared with KD II and KD III injuries.Table 2.: Percentage of patients with vascular or nerve injury for each Schenck classification categoryOverall, there is evidence that the Schenck classification system is useful in predicting vascular and nerve injuries after dislocations and stratifying the risk of poor postoperative outcomes in patients with KD IV and KD V injuries. However, lack of interobserver and intraobserver validation of the classification is a major limitation that clinicians must consider when using it in practice. Although we believe that isolated ligament tests likely remain valid in the setting of a multiple-ligament injury, the risk remains that unvalidated classification systems may badly mislead clinicians and lead to patient harm. Limitations The current limitations of the Schenck classification system stem in large part from the infrequency of knee dislocations, which limits the patient population available for study. As a result of the small patient population, the classification as a whole has not been validated, which is the most obvious limitation of the grading scheme. KD I and KD II-type injuries are especially uncommon, and predicting the extent of injuries and outcomes of patients with KD I and KD II injuries has not been extensively studied. Furthermore, the prognostic value of the classification is limited in KD I, KD II, and KD III injuries, as evidenced by the lack of difference in clinical outcomes (Lysholm and IDKC scores) between the three groups [4]. Despite these limitations, we believe the classification system may provide important information. The KD III classification appears useful in suggesting a higher likelihood of nerve or artery injury compared with KD I and KD II injuries. In addition, we believe the classification system is useful when distinguishing the outcomes of the more severe KD IV and KD V injuries from the less severe KD I, KD II, and KD III injuries. In the setting of an acute injury, the Schenck classification may be less useful than the Kennedy classification when performing reduction, as post-injury reduction technique is based on positional displacement of the tibia. For that reason, the Schenck classification is better used for surgical planning preoperatively than for joint reduction. Recommended treatment after knee dislocation often includes immediate reduction followed by later reconstruction, regardless of classification [12]. However, some patients may be treated nonoperatively after reduction, particularly less-active patient populations. For example, in low-energy knee dislocations occurring in patients with morbid obesity (who may also have more medical comorbidities that can influence surgical decision making), prolonged immobilization remains an option. As a result of controversy regarding the appropriate treatment for knee dislocations, the system is limited in its ability to provide a definitive treatment plan. We believe more research must be done to determine whether an alternative approach yields acceptable results for injuries classified as KD I or KD II, or in patients with obesity or low physical activity. However, we believe that the classification can be used preoperatively during surgical planning because patients with KD I, KD IV, and KD III injuries undergoing surgery may also undergo posteromedial or posterolateral corner reconstruction in addition to major ligament reconstruction [25, 29]. Finally, the Schenck classification is limited when describing the energy level of injuries and evaluating multi-ligament injuries associated with tibial plateau fractures. High-energy injuries have been shown to be associated with a higher likelihood of vascular injury [17], which is not addressed in the Schenck classification. Additionally, the description of fracture-dislocation injuries is limited, as all are classified as KD V. Therefore, tibial plateau fractures are not best characterized by the Schenck system and are more appropriately classified using the systems developed by Moore [18] or Schatzker [21]. Conclusions The Schenck classification is currently the most commonly used anatomic classification system for knee dislocations. It is an intuitive system that allows for anatomic preoperative planning. Furthermore, the classification has modest utility as a guide to determine prognosis and to predict neurovascular complications. However, because of the rarity of the injury and resulting small patient population, current research is limited, and further studies will be needed to characterize patient risks and outcomes based on classification. Furthermore, the lack of validation carries risks when the Schenck classification is used in clinical practice. Although confirmation of tests of isolated ligament injuries may suggest the validity of the Schenck classification overall, the risk remains that an unvalidated grading scheme may mislead clinicians and potentially cause patient harm. Despite its limitations, the Schenck classification does still have use as a communication, planning, and prognostic tool in the evaluation of potential associated neurovascular injuries for physicians and patients who have experienced knee dislocations.