1 Introduction

Acetabular fractures, accounting for 1–2% of all fractures, pose a significant challenge for orthopaedic surgeons due ¹ to their complex anatomy and potential for long-term functional limitations.^2,3 Accurate classification is crucial for guiding treatment decisions and prognostication. Among various classification systems, the Letournel and Judet (L-J) classification remains widely used due to its simplicity, reproducibility, and correlation with surgical approach and outcomes.⁴

However, the reliability of any classification system hinges on its consistency in application. Both interobserver reliability, reflecting agreement between different assessors, and intraobserver repeatability, signifying consistency within a single assessor over time, are crucial for ensuring confidence in its clinical utility.⁵

Despite the widespread adoption of the L-J classification, concerns persist regarding its interobserver reliability and intraobserver repeatability.⁶ The inherent complexity of acetabular fractures, coupled with the subjective nature of certain L-J fracture subtypes, can lead to discrepancies in classification.⁷ Furthermore, limited studies have comprehensively evaluated the reliability of the L-J classification, particularly in diverse populations and with varying levels of orthopaedic experience.⁸

Therefore, a thorough assessment of both interobserver reliability and intraobserver repeatability of the L-J classification for acetabular fractures is warranted. This study aims to address this gap by investigating the agreement between independent orthopaedic surgeons and the consistency of individual surgeons in classifying acetabular fractures using the L-J system. By employing robust methodologies and analyzing data from a representative sample of patients, we seek to provide valuable insights into the reliability of this widely used classification system and inform its effective implementation in clinical practice.

2 Methods and materials

In a prospective observational study conducted at B & B Hospital, Gwarko, Lalitpur, Nepal, we evaluated the interobserver reliability and intraobserver repeatability of the Letournel and Judet classification for acetabular fractures while also investigating the impact of CT scans on classification reliability. Over a one-year period, we recruited 45 skeletally mature patients with confirmed acetabular fractures on plain radiographs. Exclusions included polytrauma, associated fractures elsewhere, incomplete imaging, and age under 18.

AP and Judet view x-rays and CT scans of 45 patients with pelvic fractures were collected and randomly assigned numbers by the senior treating surgeon. These were already classified during surgery by the senior treating surgeon. Along with X-rays and CT scans, each observer was also provided with diagrams of the LJ classification system. Six observers with varying degrees of experience in pelvic and acetabulum fracture surgery were randomly selected for the study and divided into three groups. The first group consisted of two senior surgeons with seven years of experience; the second group consisted of two surgeons with three years of experience; and the last group had two surgeons with one year of experience. Each reviewer was not allowed to communicate with others and was asked to review solely based on X-rays first, then CT scans were provided later as a reference. After one month, the same X-rays and CT scans were provided with another random number for review again. The reviewers were blinded by their previous classification. In this way, interobserver reliability was assessed by comparing the reviews of all three reviewers, and intra-observer reliability was assessed by comparing reviews of the same X-rays and CT scans by the same observer over a span of 1 month.

Data collected included demographics, fracture type by Letournel and Judet classification on each modality (radiographs, Judet view, and CT scan), and surgeon classification for each modality. All data were entered into Microsoft Excel and analyzed using SPSS version 24.

Cohen's kappa coefficient was employed to assess the agreement level between observers and different imaging modalities. Kappa values were interpreted as follows: less than 0.00 (poor agreement), 0.01–0.20 (slight agreement), 0.21–0.40 (fair agreement), 0.41–0.60 (moderate agreement), 0.61–0.80 (substantial agreement), and greater than 0.80 (almost perfect agreement).

Power analysis, conducted using a two-proportion z-test with an alpha level of 0.05 and a desired power of 80 %, informed the sample size calculation. We further assessed inter-observer and intra-observer reliability using Cohen's kappa coefficient with 95 % confidence intervals. Finally, the effect of the CT scan on the reliability of fracture classification was evaluated by comparing the kappa values for classification based on plain radiographs and Judet view with those based on CT scans.ElementaryAssociated✓Anterior wall✓Anterior column✓Posterior wall✓Posterior column✓Transverse✓Posterior column with posterior wall✓Transverse with posterior wall✓T-shaped✓Anterior column posterior hemi transverse✓Associated Both-column

3 Results

A total of 45 skeletally mature patients with confirmed acetabular fractures were recruited over a one-year period. The mean age of participants was 36.75 years, with 86 % being male. Side involvement distribution showed 46 % affecting the left side and 54 % impacting the right side.

The data presented in Table 1 illustrates the interobserver reliability, measured by kappa (k) values, for two diagnostic modalities across three different groups of experience levels during two sessions. For the group with the most experience (>7 years), there was little difference in reliability when using plain radiographs alone between Session 1 (k = 0.73) and Session 2 (k = 0.71). However, combining plain radiographs with CT scans significantly improved reliability in Session 2 (k = 0.84) compared to Session 1 (k = 0.73). The experienced groups (>3 years and >1 year) showed a similar trend, with marginal improvements in k values from Session 1 to Session 2 when using plain radiographs alone, and more notable increases when adding CT scans. The >3 years group increased from k = 0.62 to k = 0.71, and the >1 year group from k = 0.68 to k = 0.71 in the second session when using both modalities, indicating that the addition of CT scans generally enhances interobserver reliability, particularly in more complex or challenging cases.

Table 3 summarizes intraobserver reliability across different experience levels, using kappa (k) values for plain radiographs alone versus a combination of plain radiographs and CT scans. In the most experienced group (>7 years), k values indicate substantial to almost perfect reliability, decreasing slightly with the addition of CT scans (from k = 0.73 to k = 0.71 for Observer 1 and k = 0.86 to k = 0.72 for Observer 2). For the groups with >3 years and >1 year of experience, reliability remains substantial with both modalities, though there's a marginal drop when CT scans are used (from k = 0.71 to k = 0.68 for Observer 1 in the >1 year group).(see Table 2)

Table 3 reports the percentage agreement between radiographic observations and surgical diagnoses across two sessions involving forty-five radiographs. The data, encompassing 1083 observations, reveals varied levels of agreement across different diagnoses. For instance, agreement for ABC increased slightly from 65 % to 68 % across sessions, while more substantial improvements were seen in AC, from 86 % to 89 %. Conversely, the agreement for ACPHT was lower, moving from 52 % to 56 %. High consistency was observed in PW, with agreement percentages at 96 % and 98 %, respectively. Notably, the overall agreement across all cases decreased slightly from 86 % in the first session to 83 % in both sessions combined, indicating potential issues with observer consistency or interpretation accuracy over time.

4 Discussion

Open reduction and internal fixation of displaced acetabular fractures is the preferred method of treatment to preserve the hip joint.⁹ However, the main challenge in treating these fractures lies in the limited understanding of their anatomical and pathological nature, which complicates the selection of a surgical approach for accurate reduction. This often results in unpredictable outcomes.¹⁰ The work of Letournel has been pivotal in enhancing our understanding of these fractures. He developed a classification system and a more effective management algorithm, which have become essential tools in guiding surgical treatment. Letournel's classification system is now the most widely used method for assessing and comparing clinical reports.⁸

This study evaluated the interobserver reliability and intraobserver repeatability of the Letournel and Judet (L-J) classification for acetabular fractures, examining the impact of CT scans on classification consistency among orthopaedic surgeons.Our findings align with previous studies asserting the reliability of the L-J classification system in clinical settings, emphasizing its widespread acceptance and utility in fracture diagnosis.¹¹

Interobserver reliability was quantified using kappa values, which reflected substantial agreement among observers. For the most experienced surgeons (>7 years), the interobserver kappa values were relatively stable when using only plain radiographs across two sessions (k = 0.73 and k = 0.71). The stability of interobserver kappa values among the most experienced surgeons using plain radiographs might be attributed to their extensive expertise and practice, which enhance their ability to consistently interpret radiographic features. Standardized training and familiarity with diagnostic criteria further reduce variability in assessments. Additionally, experienced surgeons exhibit confidence in decision-making and effective communication, leading to substantial and stable agreement reflected in kappa values across sessions.¹² However, the addition of CT imaging in the second session significantly enhanced the kappa value (k = 0.84), suggesting that CT scans may facilitate a clearer and more consistent interpretation of complex fractures.¹³ This enhancement was similarly observed in less experienced groups, indicating that the integration of CT imaging might reduce the subjectivity and variability inherent in radiographic fracture classification.⁸

Intraobserver reliability was also significant, although slight decreases in kappa values were noted with the addition of CT scans for some observers Plain radiographs only (k = 0.86) and with CT (k = 0.72) for observer two and similar findings in a less experienced group of more than one year and in the groups of experience more than three years similar findings have been noted. This decrease may be attributed to the increased complexity and detail provided by CT imaging, which could introduce a level of interpretation variability not present with plain radiographs alone.⁸ Despite this, the overall substantial kappa values indicate a strong consistency in individual observers' ability to classify fractures using both imaging modalities over time.¹⁴

Percentage agreement between radiographic observations and surgical diagnoses also provided insight into the practical efficacy of the L-J classification. High agreement levels in specific fracture types, such as PW, which reached up to 98 % in the second session, underscore the accuracy of radiographic assessment in clear-cut cases. However, the overall slight decline in agreement from the first to both sessions combined (86 %–83 %) highlights potential challenges in maintaining consistency across more ambiguous or complex fracture types, which may necessitate additional training or revised classification guidelines to improve.¹⁵

Ethics statement and conflict of interest disclosures

Human subjects: Consent was obtained or waived by all participants in this study. B&B Institutional Review Committee - B&B IRC issued approval B&BIRC-22-16. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue. Conflicts of interest: In compliance with the ICMJE uniform disclosure form, all authors declare the following.

CRediT authorship contribution statement

Deependra Yadav: Concept and design, Acquisition, analysis, or interpretation of data, Drafting of the manuscript, All authors have reviewed the final version to be published and agreed to be accountable for all aspects of the work. Bibek Banskota: Concept and design, Acquisition, analysis, or interpretation of data, Drafting of the manuscript, Critical review of the manuscript for important intellectual content, All authors have reviewed the final version to be published and agreed to be accountable for all aspects of the work. Rajendra Aryal: Concept and design, Acquisition, analysis, or interpretation of data, Drafting of the manuscript, Critical review of the manuscript for important intellectual content, All authors have reviewed the final version to be published and agreed to be accountable for all aspects of the work. Nitesh raj Pandey: Concept and design, Acquisition, analysis, or interpretation of data, Drafting of the manuscript. Ansul raj Bhandari: Concept and design, Acquisition, analysis, or interpretation of data, Drafting of the manuscript, Critical review of the manuscript for important intellectual content, All authors have reviewed the final version to be published and agreed to be accountable for all aspects of the work. Rajan Bhusal: Concept and design, Acquisition, analysis, or interpretation of data, Drafting of the manuscript, Critical review of the manuscript for important intellectual content, All authors have reviewed the final version to be published and agreed to be accountable for all aspects of the work. Ashok Kumar Banskota: Concept and design, Acquisition, analysis, or interpretation of data, Drafting of the manuscript, Critical review of the manuscript for important intellectual content, Supervision, All authors have reviewed the final version to be published and agreed to be accountable for all aspects of the work.

Consent of the guardians/patients

All patients, or their guardians where applicable, provided informed consent for the use of X-ray images for research purposes.

Funding statement

Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.