Background

Femoral neck fractures are common in orthopedic injuries with bimodal distribution. In different age groups, the mechanism of injuries and treatment strategies are considerably different. The young adults are more likely to suffer from high-energy injuries, compared to older individuals. Pauwels angle measurement has been utilized comprehensively as its accuracy and reliability were addressed in various researches [1,2,3]. Additionally, Pauwels classification can be used to guide the implant selection. The dynamic hip screw (DHS) is recommended for patients with Pauwels type 3, while multiple cannulated screws (MCS) are recommended for patients with Pauwels types 1 and 2 [4].

However, the fracture displacement and technical errors during X-ray as well as the variation of backgrounds and experiences of the observers diminish the accuracy and reliability of preoperative plain film evaluation in Pauwels angle measurement [5]. This gives rise to the new method of measuring Pauwels’s angles. Wang , Yang, Shen et al. described the preoperative modified Pauwels angles as an alternative to the traditional Pauwels angles [6]. Their results have shown the significantly improved intra-observer and inter-observer reliability of the modified Pauwels angles method when compared to the traditional ones. Subsequently, the modified Pauwels angle measurement is found to be more precise and reliable [5].

According to Zhang et al., preoperative and postoperative Pauwels angles were significantly different at 10.66 (± 6.47) [7]. This led to the importance of measuring methods in order to achieve the most accurate and reliable Pauwels angle measurements. Nonetheless, based on our comprehensive research, there were no data on the effective method to measure the modified Pauwels angle. Therefore, this study was aimed to examine the reliability of the modified Pauwels angle measurements both preoperatively and intraoperatively, which will eventually lead to appropriate treatment plans for patients.

Materials and Methods

Ethical approval from Maharat Nakhon Ratchasima Hospital Institutional Review Board (MNRH IRB) was obtained for this study. Sixty-six patients with femoral neck fractures who underwent fixation procedures from August 2020 to February 2022 at our hospital were included. The preoperative standard anterior–posterior (AP) and lateral radiographs as well as intraoperative fluorography were evaluated. The treatment decisions had been made based on the intraoperative fluorography.

This study included the patients aged 15–65 with a femoral neck fracture who were treated with fixation procedure. The exclusion criteria were the patients with ipsilateral peritrochanteric fractures, pathological fractures, and neglected fractures (> 30 days) [8]. The observers consisted of two trauma surgeons and five orthopedic residents.

Preoperative and Intraoperative Assessment

Seven observers, with different levels of experience, performed preoperative and intraoperative modified Pauwels angle measurements: two trauma orthopedists and five orthopedic residents with 3 to 4 years of experience.

For each radiograph or fluorography, the modified Pauwels angle was measured twice by each observer. On the part of accuracy assessment, the measurement had been done in random orders with more than 1-week interval.

For the preoperative assessments, the observers were assigned to evaluate the modified Pauwels angles and types from plain radiographs of the patients on standard AP view of both hips. The position for radiography was arranged by setting the patients in the supine position with medially rotating feet approximately 15–20 degrees, making the femoral neck to be perpendicular to the X-ray beam. The patients’ heels were placed separately with 20–25 cm of space in between. The radiographs’ quality would be validated by evaluating the lesser trochanter profile; new radiographs may be requested for the qualified radiographs. At last, the angles were measured by using PACS’ angle measurement program tool (Fig. 1).

Fig. 1
figure 1

The modified Pauwels angle measurement and typing. Use the central line of the femoral shaft as a reference, and draw a perpendicular line. The modified Pauwels angle could be measured as the angle between the perpendicular and fracture lines

Full size image

For the intraoperative assessments, patients were set in the supine position on a fracture table or radiolucent table. After reduction of the femoral neck fracture to an acceptable alignment (the most comprehensive view of the femoral neck based on the lesser trochanter profile), a fluorography was taken and printed. The angles were measured by using a goniometer (Image intensifier (fluoroscope): Siemens healthiness Cios Select 2016 model No.10893460, 10,847,833, 10,847,834).

Sixty-six femoral neck fractures were included in our study. The fractures’ angles were classified into three types: type I, less than 30 degrees; type II, between 30 to 50 degrees; and type III, more than 50 degrees.

Analysis

For intra-observer and inter-observer levels, the reliability was assessed on two outcomes: modified Pauwels angle and modified Pauwels type (Table 1). For modified Pauwels typing ordinal (measurement), we used SPSS statistics software (version 26.0) for intra-observer reliability. To calculate the multi-rater Kappa coefficient for the inter-observer agreement, the statistical method of Fleiss’ kappa was used [9]. We interpreted the kappa value according to the guideline proposed by Landis and Koch: less than 0.00 indicates poor reliability, 0.00–0.20 indicates slight reliability, 0.21–0.40 indicates fair reliability, 0.41–0.60 indicates moderate reliability, 0.61–0.80 indicates substantial agreement, and 0.81–1.00 indicates almost perfect agreement [10]. For continuous measurement (modified Pauwels angle), the intraclass correlation coefficient (ICC) was used. The two-way ICCc analysis for single rater and multiple raters was computed for intra-observer and inter-observer ICCs. Based on the 95% confidence interval of the ICC estimation, less than 0.5 showed poor reliability, 0.5–0.75 showed moderate reliability, 0.75–0.9 showed good reliability, and greater than 0.90 showed excellent reliability [11].

Table 1 Outcomes, definitions, and statistical methods applied to analyze reliability
Full size table

Results

A total of 66 patients with femoral neck fractures treated with fixation procedures (dynamic hip screw, multiple screws/cannulated screws, and cephalomedullary nail) were included. The majority of the patients are males (48 out of 66) with mean patient age of 36.95 years (SD = 14.50). Sides of the femoral neck fractures were not significantly different (30 of right vs 36 of left). The average time to surgery was 3.25 days. The modified Pauwels angles and types were assessed by seven observers independently.

Even the intra-observer reliability of modified Pauwels angle on preoperative assessment was reported with good outcomes, at ICCs: 0.804 (95% CI 0.724–0.868), the intraoperative reliability appeared to be slightly superior with excellent reliability, at ICCs: 0.943 (95% CI 0.916–0.963). The inter-observer for preoperative and intraoperative were 0.675 (95% CI 0.587–0.760) and 0.834 (95% CI 0.779–0.884) for all observers, 0.977 (95% CI 0.962–0.986) and 0.982 (95% CI 0.970–0.989) for trauma orthopedists, and 0.622 (95% CI 0.52–0.721) and 0.823 (95% CI 0.760–0.876) for residents, respectively (Chart. 1).

Chart. 1
figure 2

ICC’s value in modified Pauwels angle: intraobserver and Interobserver for all observer, specialty trauma orthopedist and residents

Full size image

The modified Pauwels typing on preoperative assessment showed moderate level of intra-observer reliability; the Kappa coefficient value was 0.584 (95% confidence interval: 0.580–0.588). On the contrary, the intra-observer reliability on intraoperative assessment showed almost perfect agreement with 0.823 (95% CI 0.819–0.828). The inter-observer reliability for preoperative and intraoperative assessment was 0.467 (95% CI 0.465–0.468) and 0.753 (95% CI 0.751–0.754) for all observers, 0.647 (95% CI 0.639–0.654) and 0.783 (95% CI 0.775–0.790) for specialty trauma orthopedists, and 0.41(95% CI 0.408–0.412) and 0.752 (95% CI 0.750–0.755) for the residents, respectively (Chart. 2).

Chart. 2
figure 3

Kappa value is modified Pauwels type: intraobserver and Interobserver for all observers, specialty trauma orthopedists, and residents

Full size image

However, there were significant differences between the modified Pauwels angles in preoperative and intraoperative measurements; the absolute difference was 9.75 ± 6.76 (range: 0 to 30.5 p < 0.05). There was disagreement on modified Pauwels type between the preoperative and intraoperative evaluations in 11 patients (16.67%) (Fig. 2). When discordance occurred, the final decisions depended on intraoperative fluorography. The implant selection was changed based on Pauwels type in eight patients (12.12%).

Fig. 2
figure 4

An example of an assessment of modified Pauwels angle and type comparing preoperative A and intraoperative B reviews. There were different angles, and the modified Pauwels type rating was changed from type 3 to type 2

Full size image

Discussion

Femoral neck fractures are common among orthopedic trauma cases. The preceding studies have assessed the value of modified Pauwels angles as a risk factor [12] for loss of reduction and fracture nonunion or malunion in the postoperative follow-up. It is also reported as a risk factor for subsequent avascular necrosis [1, 13]. In the previous studies, the Pauwels type III cases with excellent reduction and proper implant placement were reported with nonunion rates of 19% for fractures treated with cannulated screws alone and 8% for those treated with a fixed-angle device (such as DHS) [14]. Therefore, it is essential to measure the modified Pauwels angle with the most accurate and reliable method in order to guide the appropriate implant selection with the least complication. A comparison of DHS and MCS surgeries revealed that the patients who underwent MCS fixation had shorter hospital stays, lower rate of acute myocardial infarction and pneumonia after 30 days, lower rate of mortality within 90 days, lower rate of avascular necrosis, a significantly smaller incision, and less intraoperative blood loss when compared to the DHS group [6, 15]. Therefore, MCS is a preferable option if the fractures are the lower types of Pauwels classification.

On modified Pauwels type, the preoperative assessment had moderate intra-observer and inter-observer reliability in all observers and all subgroups. On the other hand, higher reliability was noted in the intraoperative assessment of modified Pauwels types with excellent level of intra-observer reliability and good level of inter-observer reliability (all observers).

For the preoperative modified Pauwels angle assessments, the results demonstrated good intra-observer reliability. Inter-observer reliability of preoperative assessment showed moderate reliability in all groups, except that excellent reliability was noted in trauma orthopedist group. Nonetheless, the intraoperative assessment had higher intra-observer and inter-observer reliability, regardless of the observer groups. We reported excellent reliability among the trauma orthopedists, and good reliability among the others. However, the reliability also depended on the variability of patient positioning, pre-reduction displacement, radiographs quality, and observers’ experiences. This study reported higher reliability of intraoperative evaluation due to the ability of fluorographys’ quality control. In addition, the observers’ experience and capability had significantly affected the reliability of radiographs interpretation. Among all observers, specialized trauma orthopedists had the highest reliability in interpreting both preoperative and intraoperative modified Pauwels types and angles.

As demonstrated in this study, surgical implants were determined by the modified Pauwels type. The absolute difference of modified Pauwels angle in preoperative and intraoperative assessment was 9.75 ± 6.76 degrees (range: 0 to 30.5 p < 0.05). In concordance with Zhang Y.L [7]. analysis, their result also showed significant differences in Pauwels angle between preoperative and postoperative radiographs (P = 0.037), with the absolute difference of 10.66 ± 6.47 (range 1.72–38.48). This amount of angle differences could alter the Pauwels type as well as implant adjustment. Therefore, the intraoperative Pauwels angle measure should be used to determine the implant and fixation method [16].

Even though we had strictly followed the protocols in radiography and fluorography technique, the minimal technical errors could occur and affect the interpretation accuracy. Another limitation of this study was that the timing between the preoperative and intraoperative measurements had not been recorded in this study. Further study should focus on the use of new instruments such as the femoral neck system [17] with the report of midterm to long-term postoperative clinical outcomes after fixation.

Conclusion

Our study proposed acceptable reproducibility of the modified Pauwels angle intraoperatively and excellent reliability to preoperative plain radiographs. Due to its high reliability and reproducibility, modified Pauwels angle and type is an effective tool for initial assessment and surgical planning. The intraoperative modified Pauwels angle and type assessment offered excellent accuracy and appropriate implant selection. Furthermore, two or more specialized trauma orthopedist’s agreement on evaluation resulted in maximum reliability and accuracy. Finally, the authors recommended the use of an intraoperative modified Pauwels angle and type assessment to accurately determine implant selection for femoral neck fracture fixation.