The oblique plane deformity in slipped capital femoral epiphysis

The reliability for our method of measuring the deformity in the coronal and axial planes can be classified as almost perfect agreement [10], with ICC values of 0.964 and 0.914, respectively. Our method of combining these two measurements for each patient has also been shown to have near perfect agreement (ICC 0.947) in terms of calculating the true magnitude of the slip in the 20 cases examined. This suggests that it could be reliably repeated in clinical practice in a more widespread manner. In our series, there was a wide variety of orthopaedic experience among the four assessors, which makes the reliability even more clinically relevant. In addition, there was substantial to almost prefect agreement in the intraobserver reliability for the test re-test results, despite a gap of 6 months between the two tests and no further training being given to the raters.

An AP radiograph in the supine position is easy to obtain and gives a reliable view. There are, however, certain technical difficulties associated with a lateral radiograph of the hip, particularly in a child with a painful SCFE. The standard frog lateral position can be painful and risks exacerbating the initial external rotation deformity [4]. The method described by Billing requires the femur to be externally rotated 90°, elevated 25° from the table, with the knee flexed 90°. This has been shown to be an accurate method of diagnosing a minor (particularly contralateral) slip, but has the same drawbacks as the frog lateral position for patients with possible major slips [11, 12]. In addition, the reliability of the Billing method has been questioned in cadaveric studies [13]. Furthermore, the more severe the deformity, the more restricted hip flexion and abduction becomes; thus, standardised lateral views become impossible. Other methods of obtaining a lateral view such as shoot-through or cross-table can be associated with significant soft-tissue shadow [14] (particularly in patients with a large body mass index, as is increasingly seen in these patients [15]) and may, therefore, be difficult to interpret [16]. For the purposes of this study, none of these “lateral” views can be guaranteed to give a truly orthogonal view to the AP radiograph, which is an essential prerequisite for using the geometrical method to calculate the magnitude of the oblique plane deformity [8].

These problems are overcome with cross-sectional imaging, such as CT or MRI. Both methods produce a true orthogonal view in the form of an axial image, which allows measurement of the posterior angulation and torsional component of the deformity. By combining this measurement with the displacement on the AP radiograph, the magnitude and direction of the true deformity in the oblique plane can be calculated. In our unit, we investigate children referred with possible moderate or major SCFE with both the AP radiograph and either or both of these methods. MRI was chosen in certain patients in order to assess for possible pre-operative avascular necrosis (AVN) [17]. For the purpose of measuring the slip angle, we do not believe that there is a significant difference between the two methods. However, the advantages of CT scanning are that it is more readily available, quicker to perform and has a lower cost than MRI scanning. However, the drawbacks are that the lifetime additional radiation risk of malignancy is increased from 1 in 100,000 for a pelvis X-ray to 1 in 10,000 for CT, thus increasing the risk from “very low risk” to “low risk”. This should be taken in the context of a 1 in 3 population risk of developing malignancy [18]. Furthermore, the radiology department at our institution is experienced in performing cross-sectional imaging for the assessment of rotational profiles. As such, it is able to provide the required cross-sectional imaging at the hip using targeted CT to minimise the radiation dose. Standard MRI sequences have not been found to be useful predictors of AVN in SCFE; however, the use of digital subtraction sequences has been shown to be promising in preliminary studies [9]. The advantage of determining whether there is pre-existing AVN is that it allows for greater confidence in predicting prognosis and planning further treatment.

Having calculated the oblique plane deformity, the management plan can then be made with more confidence. If the true magnitude is higher than first appreciated on initial radiographs, the decision may be made to perform open reduction or a corrective osteotomy. In addition, the degree of the true magnitude in the oblique plane may be used as a guide to prognosis.

We chose to use the AP pelvis rather than a coronal CT reconstruction to measure the deformity in the coronal plane because of our aim to identify a method which is readily available. The rationale for this was that some authors recommend urgent fixation of acute slips [4] and, as such, although it is often possible to obtain CT scans outside of normal hours, it is not always possible in our institution to obtain coronal re-formats. As such, we wished to evaluate whether our method was applicable using the readily available AP pelvis X-ray in order to provide a junior registrar with the means to rapidly assess whether a slip was suitable or not for in situ pinning.

There are several limitations to our study. The number of cases was relatively small; however, given that they all represented moderate or severe SCFE, there was not a significant heterogeneity of values in our samples.

Although it was not part of the original aims of the study, we found that several patients were seen on CT or MRI to have a marked curve to the femoral neck itself, rather than a straight neck and a sharp deformity at the physis (Fig. 3). These represented chronic slips with re-modelling. This is more in keeping with one of the early descriptions of SCFE by Ernst Muller in 1888 as “Schenkelhalsverbiegungen im Jungesalter”, meaning “bending of the femoral neck in adolescence” [19]. Knowledge of this deformity would be of obvious benefit before embarking on surgical intervention, as measurement of the displacement of the physis may be more difficult as re-modelling of the femoral head and neck occurs. Although re-modelling in chronic SCFE can cause retroflexion of the neck, the points on the femoral neck and head to measure the slip angle can be accurately delineated.

The magnitude of the mean angles in our series also gives important information. The mean deformity (66°, range 43–83) in the axial plane was twice that of the mean coronal plane deformity (33°, range 4–63). The mean oblique plane deformity was even higher at 75° (range 43–98). This shows that the true deformity can be significantly underestimated when relying on one image, particularly the AP radiograph. This study demonstrates that the true magnitude of the deformity in the oblique plane can be accurately and reliably measured with little inter- and intraobserver variability.