Langenskiöld Classification for Blount Disease: Is It Reliable?

BACKGROUNDS: Blount disease (BD) is a developmental disease which medial part of the proximal tibial physis resulting genu varum. Plain radiographs are necessary for diagnosis. For this purpose, Langenskiöld classification is used. There are not enough data about Langenskiöld classification's reliability in the English literature. The purpose of this study is to evaluate the intra- and inter-reliability of Langenskiöld classification in BD.
MATERIALS AND METHODS: Thirty seven patients with BD were included in this cross-sectional study. Two separate presentations including clinical findings and standard knee radiographs were evaluated by three different groups (residents, surgeons, and pediatric orthopedic surgeons) and were sent to the observers every other month by hiding patients' personal information. Intraobserver and interobserver reliabilities in these three groups were examined. Intraclass correlation coefficient (ICC) tests were used for this purpose.
RESULTS: Intraobserver reliability (ICC) was similar (excellent) in three groups (0.822, 0.804, and 0.763). Interobserver reliability among three groups was meanly excellent.
CONCLUSION: Agreement to Langenskiöld classifications found excellent for three groups. Surgical experience did not change intraobserver and interobserver reliability in Langenskiöld classification.

Introduction

Blount disease (BD) is a developmental disease which is characterized by abnormality of endochondral ossification of medial part of the proximal tibial physis.1 As a result, a genu varum deformity occurs. At the following years, a progressive gait abnormality, limb length discrepancy, and also early osteoarthritis of the knee can be observed.234

A plain radiography is the main tool for diagnosis.5 Clinical classifications were developed as early or late onset which uses 4 years of age as a threshold, and at the following years, late-onset form was further divided as juvenile and adolescent types.678 Radiological classification was introduced by Langenskiöld.7 One can expect reversible changes in Type 1 through Type 3.9 Types 5 and 6 are known as high-grade disease which is difficult to be treated.10 Langenskiöld classification system is useful. There are not enough data about Langenskiöld classification's reliability in the current literature,15 so we aimed to evaluate the intra- and inter-reliability of Langenskiöld classification.

Materials and Methods

Thirty-seven patients with BD were included this cross-sectional descriptive study after the approval by the ethics committee. Hospital's picture archiving and communication systems were used to assess patients’ standard knee joint anterior–posterior (A-P) and lateral radiographs retrospectively. These 37 patients who had unilateral knee involvement and no additional diseases were included in the study. Two sets of Microsoft® PowerPoint Presentations including each patient's radiographs as a single slide were prepared in which the order of cases was randomized and blinded [Figure 1]. Langenskiöld classification was presented to the observers after visual presentation of each patient.

Figure 1

Presentation of radiographs of the patient assigned as Case no 1 in Set 1

Observers were divided into three groups based on their surgical experience. In the first group, there were ten residents trained in the Department of Orthopedics and Traumatology at the training and research hospital for at least 4 years. The second group included ten seniors working in a general orthopedic department for at least 3–5 years. Ten attending consultant pediatric orthopedic surgeons from different institutions with at least 5 years of surgical experiences took part in the third group. Assessments of the sets were performed in random order by each observer on two separate occasions, at least 4 weeks apart. Examples of the user interface typical of each case are shown in Figure 1. Observers were asked to report the stage which was most appropriate for each patient.

Statistical analysis

Statistics were analyzed using the Statistical Package for the Social Sciences (SPSS) version 21 standard statistical version (SPSS Inc., Chicago, IL, USA). Interobserver and intraobserver reliabilities were calculated using intraclass correlation coefficient (ICC), w2, and P values within 95% confidence intervals.11 An ICC value greater than 0.75 is excellent, from 0.75 to 0.60 is good, from 0.59 to 0.40 is fair, and below 0.40 is poor agreement.12

Results

The mean intraobserver reliabilities between sets are given in Table 1. Interobserver reliabilities of the each group are shown in Table 2. In both assessments, strengths of ICC values were excellent for all groups but ICC values decreased in the second evaluation.

Table 1

Mean intraobserver reliability values

Groups	ICC	95% CI (minimum-maximum)	Strength of reliability
Residents (n=10)	0.822	0.780-0.856	Excellent
Seniors (n=10)	0.804	0.758-0.842	Excellent
Pediatric orthopedists (n=10)	0.763	0.707-0.809	Excellent

ICC=Intraclass correlation coefficients, CI=Confidence interval

Table 2

Interobserver reliability values

Groups	Set I			Set II
	ICC	95% CI (minimum-maximum)	Comment	ICC	95% CI (minimum-maximum)	Comment
Residents (n=10)	0.943	0.909-0.968	Excellent	0.888	0.822-0.937	Excellent
Seniors (n=10)	0.930	0.888-0.960	Excellent	0.928	0.885-0.959	Excellent
Pediatric orthopedists (n=10)	0.931	0.890-0.961	Excellent	0.929	0.887-0.960	Excellent

ICC=Intraclass correlation coefficients, CI=Confidence interval

In addition, intra- and interobserver reliabilities were not found to be affected by the level of experience [Tables 1 and 2].

Discussion

Langenskiöld classification system has been used to stage infantile tibia vara (BD) for many years. This classification is useful for selecting appropriate treatment modality. It is crucial to make difference between Stage 3 and Stages 4–6 to predict the prognosis.1 Although this classification system is of importance, there are not enough data about its reliability. This current study showed that both intra- and interobserver reliabilities of Langenskiöld classification were excellent and being experienced did not affect the reliability positively.

The reliability of Langenskiöld classification was evaluated first by Stricker et al. in 1994.5 The authors stated that ICC values of interobserver reliability about this classification system were 0.84 and 0.83 for pediatric orthopedic specialists and senior residents, respectively. Interobserver reliability of Stages 1 and 6 was found to be higher than the intermediate stages. In our study, group ICC values of interobserver reliability were 0.94 and 0.93 for residents and pediatric orthopedic specialists, respectively. Our results also showed that being experienced did not affect the reliability positively like the aforementioned study. We did not evaluate the reliability separately for each stage, so we cannot state that reliability is lesser for intermediate stages. One additional interesting statement of Stricker et al.'s study was that the reliability of Langenskiöld classification system decreased after modifying the classification system to four stages.

In a recent study, LaMont et al. questioned the accuracy of Langenskiöld classification.13 The authors drew attention to the high recurrence rates of Stages 2 and 3 disease and offered a new three-staged (Types A, B, and C) classification system based on the morphology of the metaphyseal/epiphyseal slope which they believed to be more correlated with treatment outcomes than the traditional Langenskiöld classification.13 The authors reported recurrence rates as 22.5%, 20.7%, and 71.7% for Stages A, B, and C, respectively. Age of 5 years was accepted as a critical transition point about the success of simple high tibial osteotomy to realign the extremity. Inter- or intraobserver reliability of this new classification was not evaluated and accordingly could not be compared with Langenskiöld classification. It should be more informative if these new classification systems and traditional Langenskiöld classification's reliabilities can be compared.

The reliabilities of each stage were not evaluated in this current study; this can be accepted as a limitation. In the literature, the only study in which Langenskiöld classification was made in more detail belongs to Stricker et al.5

Conclusion

Our study showed that both intra- and interobserver reliabilities of Langenskiöld classification were excellent, and the reliability was not associated with the level of experience.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form the patients have given their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.