Reliability of Lichtman s classification for Kienböck s disease in 99 subjects

Size: px

Start display at page:

Download "Reliability of Lichtman s classification for Kienböck s disease in 99 subjects"

Clara Beasley
6 years ago
Views:

1 Reliability of Lichtman s classification for Kienböck s disease in subjects Masaki Shin, M.D., Masahiro Tatebe, M.D., Hitoshi Hirata, M.D., Shukuki Koh, M.D., Takaaki Shinohara, M.D. Department of Hand Surgery, Nagoya University School of Medicine, 1 1 Corresponding author Masahiro Tatebe Department of Hand Surgery, Nagoya University School of Medicine, Tsurumaicho, Showaku, Nagoya -0, Japan Tel. -1-1, Fax tatebe@med.nagoya-u.ac.jp This study is a retrospective study and the approval was given by the institutional review board (IRB) Key words: Kienböck s disease, wrist, classification 0 1

2 Reliability of Lichtman s classification for Kienböck s disease in subjects Abstract Purpose: The objective of this research was to investigate the reliability of Lichtman s classification for Kienböck s disease. Methods: Interobserver reliability and intraobserver reproducibility were investigated by interpreting both anteroposterior and lateral X-rays of the wrist joint twice in patients with Kienböck s disease using the modified Lichtman s classification system. Observers comprised orthopedic surgeons, and no information was exchanged between observers either before or during the study. Results: Intraobserver reliability was moderate (0.1-0.), and interobserver reliability was fair (Siegel s kappa=0.). Conclusion: Low values were obtained regarding interobserver reliability for the modified Lichtman s classification of Kienböck s disease. This classification is thus inadequate for use in clinical settings. A new classification should be established. 1

3 Introduction Lichtman s classification is frequently used in the assessment and management of Kienböck s disease (1)(). Radiological assessment, however, is not an easy task and all classification systems used in Kienböck s disease show some degree of inter- and intraobserver variability. In the past, isolated studies have examined the reliability of Lichtman s classification of Kienböck s disease, but sample sizes in those reports have been relatively small (,, and cases) ()()(). The purpose of the present study was to assess the reliability of this commonly used classification system, and to determine possible reasons for low reliability associated with Lichtman s classification Materials and methods We reviewed radiographs of all patients with Kienböck s disease who were treated in our hospital from 1 to 00. The inclusion criterion was the availability of complete preoperative radiographs of anteroposterior and lateral views bilaterally. Bilateral cases and cases diagnosed based on magnetic resonance imaging alone were excluded. All radiographs of the wrist were obtained with the shoulder in 0 of abduction, the elbow in 0 of flexion, the forearm in neutral rotation, and the wrist in neutral alignment, utilizing a wrist support. We included cases in the present study. The observers comprised orthopedic surgeons with 1, 1, and years of experience. These observers assessed Lichtman s classification of radiographs in a blinded manner on two separate occasions. We used the modified Lichtman s classification system, consisting of stages, including

4 a subdivision of stage into stage a (no carpal collapse) and stage b (carpal collapse and fixed scaphoid flexion) (Fig. 1). No information was exchanged between observers either before or during the study. Before carrying out assessments, each observer was asked to read the original article on the classification system. We then determined the inter- and intraobserver reliabilities of the classification system. All study protocols were approved by the institutional review board of our institute. 1 1 Statistics Kappa statistics were used for the assessment of inter- and intraobserver reliabilities of the modified Lichtman s classification. We also examined inter- and intraobserver correlations between each of the individual radiological parameters using kappa values. According to Landis and Koch (), a kappa value of is considered fair, moderate, substantial and excellent Results Results for intraobserver reproducibility are summarized in Tables 1 and. For the modified Lichtman s classification, kappa values ranged from 0.1 to 0., indicating moderate agreement. Interobserver reliability for the modified Lichtman s classification showed a kappa value of 0. (Siegel s kappa; % confidence interval (CI): ) (Table ). The classifications thus showed fair agreement. Intra- and interobserver agreement for stages a and b were.% and.%, respectively. Interviews after the examination revealed that the three examiners determined the

5 classifications of stage a and b by referring to the contralateral side Discussion Radiography is the most commonly performed investigation when assessing the severity of Kienböck s disease. In 1, Lichtman proposed a classification system to assess the severity of this disease (1). The original classification was based on anteroposterior radiographs. Lichtman s classification is well-accepted, with most reports depending on this classification ()(), while clinical results have shown no correlation with radiographic stage ()(). For the original Lichtman s classification, Jensen et al. reported poor reliability from a study of cases () and Jafarnia et al. reported good reliability from a study of cases (). For modified Lichtman s classification, Goldfarb et al. reported reliability/reproducibility in cases using an additional criterion (subdividing stage radioscaphoid angle based on a cutoff of 0 ) (). Sample sizes in those reports were relatively small. The most important aspect of the present study was the thorough and systematic analysis of the reliability of Lichtman s classification systems in a relatively large sample of patients with Kienböck s disease. An ideal radiological classification system should aid in assessing the severity of the disease and help in deciding on treatment and prognosis. Furthermore, the classification system should offer reasonable inter- and intraobserver reliabilities and should not use obscure measurements. While various classifications have been proposed in the past to achieve these goals, no ideal classification system for

6 Kienböck s disease has yet been established (1)()(1). The present study included complete radiographs from patients with Kienböck s disease, and the reliabilities of the classification systems for this disease were assessed. The results showed that the modified Lichtman s classification offers only fair interobserver reliability even with experienced orthopedic surgeons. According to the current results, one cause of disagreement arises from the identification of stages a and b. Abe et al. recently reported that a cortical ring sign indicates only scaphoid flexion (1). All three observers in the present study referred to the contralateral side to determine the classification of stages a and b, but these stages do not indicate a pathological condition and the cutoff between a and b thus has little clinical relevance. The controversy resides in stage a and b for both radiological and pathological status, which might be a reason for the lack of a correlation with clinical results. Moreover, our findings revealed low interobserver reliability in all stages. For differentiating between stages and, computed tomography (CT) may be useful to evaluate degenerative joint changes. However, we did not use CT in the present study, which may partly explain the poor reliability. We do not recommend using the modified Lichtman s classification system to aid in treatment decision-making. Most surgeons base treatment plans on the results of CT and MRI and any new classification system should include findings from plain radiography, CT, and MRI. 0 1 Conclusion The modified Lichtman s classifications showed only fair agreement even for

7 orthopedic experts. A new classification system should be established, based not only on plain radiography, but also on CT and MRI.

8 References 1. Lichtman DM, Mack GR, MacDonald RI, Gunther SF, Wilson JN. Kienbock's disease: the role of silicone replacement arthroplasty. J Bone Joint Surg 1; A: -0.. Alexander AH, Turner MA, Alexander CE, Lichtman DM Lunate silicone replacement arthroplasty in Kienböck's disease: a long-term follow-up. J Hand Surg. ;1A :01-.. Goldfarb CA, Hsu J, Gelberman RH, Boyer MI. The Lichtman classification for Kienbock's disease: an assessment of reliability. J Hand Surg. 00; A : -0.. Jafarnia K, Collins ED, Kohl HW rd, Bennett JB, Ilahi OA. Reliability of the Lichtman classification of Kienbock's disease. J Hand Surg. 000; A: -.. Jensen CH, Thomsen K, Holst-Nielsen F. Radiographic staging of Kienbock's disease. Poor reproducibility of Stahl's and Lichtman's staging systems. Acta Orthop Scand. 1; : -.. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1; : 1-.. Allan CH, Joshi A, Lichtman DM. Kienbock's disease: diagnosis and treatment. J Am Acad Orthop Surg. 001; : 1-.. Beredjiklian PK. Kienböck's disease. J Hand Surg. 00; A: 1-.. Nakamura R, Imaeda T, Miura T. Radial shortening for Kienbock's disease: factors affecting the operative result. J Hand Surg ; 1B:0-.. Iwasaki N, Minami A, Oizumi N, Yamane S, Suenaga N, Kato H.Predictors of

9 clinical results of radial osteotomies for Kienbock's disease. Clin Orthop Relat Res. 00; 1:1-.. Stahl F. On lunatomalacia (Kienbock s disease). A clinical and roentgenological study, espwcially on its pathogenesis and the late results of immobilization treatment. Acta Chir Scand 1; 1 (suppl): Decoulx P, Marchand M, Minet P, Razemon JP. Kienbock's disease in miners; clinical & pathogenic study with analysis of x-rays of the wrist. Lille Chir. 1; 1: -1. French 1. Abe Y, Doi K, Hattori Y. The clinical significance of the scaphoid cortical ring sign: a study of normal wrist X-rays. J Hand Surg Eur Vol. 00; : 1-.

10 Figure legend Figure 1. Lichtman s classification. A: stage 1, B: stage, C: stage a, D: stage b, E: stage.

11 Table 1. Modified Lichtman s classification; kappa values for intraobserver reliability Observer No. observed % agreement Kappa value (%CI) (0.1-0.). 0. (0.-0.1). 0.0 (0.-0.) Years of experiences observer No. 1: 1 year, No.: 1 years, No.: years.

12 Table. Intraobserver classification Observer 1 nd examination total Stage a b 1 st examination a b total Observer nd examination total Stage a b 1 st examination 0 0 a 0 b total Observer nd examination total Stage a b 1 st examination a 0 b total 0 1 1

13 Table. interobserver classification Observer 1 and Observer total Stage a b Observer 1 a 1 0 b total Observer and Observer total Stage a b Observer a b total Observer 1 and Observer total Stage a b Observer 1 a 1 1 b total 0 1 1

ORIGINAL PAPER. Department of Hand Surgery, Nagoya University School of Medicine ABSTRACT

ORIGINAL PAPER. Department of Hand Surgery, Nagoya University School of Medicine ABSTRACT Nagoya J. Med. Sci. 74. 167 ~ 171 2012 ORIGINAL PAPER TILT OF THE RADIUS FROM FOREARM ROTATIONAL AXIS RELIABLY PREDICTS ROTATIONAL IMPROVEMENT AFTER CORRECTIVE OSTEOTOMY FOR MALUNITED FOREARM FRACTURES