Boehler’s angle and the critical angle of Gissane are of limited use in diagnosing calcaneus fractures

in the ED^B

Jason R. Knight MD^a,*, Eric A. Gross MD^a, Gail H. Bradley MD^a,

Curt Bay PhD^b, Frank LoVecchio MD^a

^aDepartment of Emergency Medicine, Maricopa Medical Center, Phoenix, AZ 85008, USA

^bDepartment of Academic Affairs, Maricopa Medical Center, Phoenix, AZ 85008, USA

Received 17 November 2005; revised 15 December 2005; accepted 17 December 2005

Abstract

Study objectives: The aim of this study was to determine the use of Boehler’s angle (BA) and the critical angle of Gissane (CAG) in diagnosing calcaneus fractures in the ED.

Design: The study was conducted as a randomized, blinded, case-control trial.

Cases: One hundred thirty-three patients older than 15 years were included in the study. Sixty-five patients with computed tomography-verified calcaneus fractures (gold standard) and 68 ED patients with lateral foot or ankle x-rays without calcaneus fractures were included in the study.

Methods: One second-year emergency medicine resident, 1 third-year emergency medicine resident, 2 board-certified emergency medicine attending physicians, and 1 board-certified radiologist prospec- tively reviewed all films using the Picture Archival and Communication System digital radiology system. Cases and controls were randomized and the participants were blinded to final radiographic diagnoses. Participants determined whether there was a fracture on each x-ray and measured BA and the CAG using the digital angle tool in the Picture Archival and Communication System.

Results: Emergency physicians were 97.9% accurate in diagnosing calcaneus fractures (range, 97% to 99%). The mean j value for emergency physicians was 0.96 (range, 0.94-0.985). Receiver operating characteristic curves were constructed for BA and the CAG. When compared with the gold standard, the area under the curve for BA ranged from 0.82 to 0.88. The area under the curve for the CAG ranged from

0.45 to 0.67. BA had an interclass correlation coefficient of 0.84 (95% confidence interval, 0.79-0.87). The CAG interclass correlation was 0.52 (95% confidence interval, 0.43-0.60). One fracture was missed by the radiologist and all of the emergency physicians because it was only visible on computed tomography. Conclusion: BA is somewhat helpful and the CAG is not useful in diagnosing calcaneus fractures in the ED. Interrater reliability for BA is excellent, but for the CAG, it is poor. Emergency physicians were 97.9% accurate in making the diagnosis by reviewing the plain films without bassistanceQ of the angle measurements. D 2006

Presented at the SAEM National Meeting, New York, NY, May 2005.

^B This study was supported by Maricopa Medical Foundation resident research grant #04-L.

* Corresponding author. Tel.: +1 602 344 5808; fax: +1 602 344 1423.

E-mail address: [email protected] (J.R. Knight).

0735-6757/$ – see front matter D 2006 doi:10.1016/j.ajem.2005.12.013

Introduction

Calcaneus fractures are frequently encountered by emergency physicians. The calcaneus is the most commonly fractured tarsal bone and accounts for 1% to 2% of all fractures [1]. In 1931, Boehler stated that measurement of the btuber-joint angleQ is a useful radiographic tool to aid in the diagnosis of calcaneus fractures [2]. He described a normal angle as 308 to 358. A Boehler’s angle (BA) less than 308 to 358 was indicative of a calcaneal compression fracture. Since that time, BA has been used to diagnose calcaneus fractures on lateral foot and ankle radiographs.

Boehler’s angle is formed by two intersecting lines: (1) a line connecting the posterior tuberosity of the calcaneus and the apex of the posterior facet of the calcaneus and (2) a line between the apex of the posterior facet and the apex of the anterior process of the calcaneus (Fig. 1) [2-4]. Unfortunately, discrepancy exists regarding a normal BA measurement. Normal values have been reported in textbooks as 208 to 408 [1,5], 288 to 408 [6], 208 to 448 [7], and 258 to 458 [8].

Another angle that has been used to diagnose calcaneus fractures is the critical angle of Gissane (CAG). The CAG is formed by two strong cortical struts that extend laterally and form an obtuse angle directly inferior to the lateral process of the talus [3,4,9-11]. The first strut extends along the lateral border of the posterior facet, and the second extends anterior to the beak of the calcaneus (Fig. 2). A normal CAG has been reported to be 1008 to 1308 [10] and 1208 to 1458 [11]. A more acute angle is indicative of a fracture, but an exact value or cutoff has not been reported in the literature. The purpose of this study was to determine the diagnostic use of BA and the CAG in patients with and without

calcaneus fractures.

Methods

Study design and setting

This case-control study enrolled 65 patients with computed tomography (CT)-verified calcaneus fractures

Fig. 1 Bohler’s angle.

Fig. 2 Critical angle of Gissane.

who presented to an urban ED between January 2000 and April 2004. The study was approved by the hospital’s institutional review board. Patients with calcaneus fractures were identified using International Classification of Dis- eases, Ninth Edition discharge Diagnosis codes. All patients had a lateral foot or ankle radiograph taken. Sixty-eight patients with lateral foot or ankle radiographs who had normal or Alternative diagnoses were used as controls.

Standard mediolateral foot and ankle radiographic projections were used. Patients were excluded if they were younger than 15 years or still had open growth plates. Patients were also excluded if they had a calcaneus fracture but did not have a CT.

Sixty-eight consecutive patients with CT-verified calca- neus fractures were identified. Sixty-five patients with fractures were included in the study. One patient was excluded because he was a multisystem trauma patient and only had a CT of the ankle performed. One patient did not have a lateral foot or ankle radiograph archived in our system, and one patient had a known fracture that was originally diagnosed 5 months before at another institution and the fracture was healing. Sixty-eight consecutive patients with normal or alternative diagnoses on plain films were selected as controls, based on order of entry into the computerized radiology system. Patients were excluded from the control group if they fell from a height.

Selection of participants

One board-certified radiologist, 2 emergency medicine residents, and 2 board-certified emergency medicine attend- ing physicians who were blinded to the patient’s final diagnosis reviewed the radiographs of all patients using a computerized digital radiology system, the Picture Archive Communication System (PACS).

Method of measurement

Readers first determined whether there was a fracture based on reviewing the plain films, and recorded their results. They then measured BA and the CAG.

Fig. 3 Receiver operating characteristic curve: Boehler’s angle.

Readers measured BA and the CAG on the PACS system using the angle tool. A guideline and pictorial on how to measure BA and the CAG was available for readers on the PACS workstation desk. Data were recorded on a standard- ized form. The order in which the x-rays were presented was randomized, and no identifying information was present.

Data collection and processing

Data were entered into a standardized data collection form by the readers as they reviewed the radiographs. Subsequently, data were entered into a formatted Excel spreadsheet. They were then imported into SPSS (version 10.1, SPSS Inc, Chicago, Ill), checked, and cleaned.

Outcome measures

The accuracy of emergency physicians and the radiolo- gist in diagnosing calcaneus fractures was determined based on comparison to the gold standard (CT). Measured angles were plotted against CT findings on receiver operating characteristic curves. If an angle had an area under the curve (AUC) of 0.9 or greater, the angle was consid- ered useful.

Data analysis

For each reader, j values were calculated against the CT. Intraclass Correlation coefficients (two-way random effects, consistency model) were used to estimate interrater reliabil- ity. An intraclass correlation coefficient of 0.8 or greater was considered bgood.Q Sixty-five subjects per group were required to calculate 95% confidence intervals (CIs) for an estimated sensitivity and specificity of 0.80 F 0.10.

Results

Emergency physicians were 97.9% accurate in diagnos- ing calcaneus fractures based on reviewing the plain films

without assistance from the angles (range, 97% to 99%). Their mean j value, when measured against the gold standard, was 0.96 (range, 0.94-0.98).

The second-year emergency medicine resident and the

2 attending emergency medicine physicians missed an avulsion fracture of the posterior lateral aspect of the calcaneus that was present on only one plain-film view (oblique). The second-year emergency medicine resident also missed a fracture through the posterior tubercle. The radiologist had a sensitivity of 98.5% and a specificity of 100%. One fracture was missed by all readers because it was only visible on CT.

Receiver operating characteristic curves were con- structed for BA and the CAG. A common cutoff for the area under the curve (AUC) in determining the usefulness of a test is 0.9. When compared with the gold standard, the AUC for BA ranged from 0.82 to 0.88 (Fig. 3). BA had an intraclass correlation of 0.84 (95% CI, 0.79-0.87). Seventeen BA measurements were negative because of severe compression.

The AUC for the CAG ranged from 0.45 to 0.67 (Fig. 4). The CAG intraclass correlation was 0.52 (95% CI, 0.43- 0.60). Several x-rays had such distorted bony anatomy that it was difficult to measure the CAG.

A subgroup analysis of patients with Compression fractures was performed because BA and the CAG are disrupted primarily in compression type injuries. The AUC for BA improved minimally (range, 0.85-0.91), and the AUC for the CAG was worse (range, 0.32-0.54) in this subgroup of patients.

Limitations

The main limitation of the study is that patients in the control group did not receive CT scans. It is possible that a patient in the control group had a calcaneus fracture not seen

Fig. 4 Receiver operating characteristic curve: critical angle of Gissane.

on plain radiographs. This would decrease the sensitivity for both BA and the CAG. Readers in our study were looking specifically for calcaneus fractures and may have a higher sensitivity of identifying fractures compared to physicians practicing in a general ED examining a radiograph obtained for bfoot painQ or other generalized complaint. The real- world sensitivity of identifying fractures may be lower because of this limitation. Finally, as we note in the Results section, BA and the CAG are disrupted primarily in compression-type injuries. Perhaps, if the study were done examining only these types of injuries, the sensitivities of the angles would be greater.

Discussion

The most common mechanism of injury leading to fractures of the calcaneus are falls from heights and motor vehicle accidents. The position of the foot at the time of impact, the force of the impact, and the quality of bone all determine the location of fracture lines and the pattern of comminution.

The downward force of the patient’s body is transmitted through the talus, which acts as a wedge and splits the calcaneus in its midportion. This downward force com- presses the posterior facet and crushes the central portion of the calcaneus. This mechanism of injury disrupts BA and the CAG.

Compression of the calcaneus results in a decrease in BA and straightening of the lines connecting the apex of the posterior facet and the superior aspect of the calcaneus, and the posterior tuberosity and the superior aspect of the calcaneus. Likewise, the CAG is disrupted when the calcaneus is crushed. The direction of the force and the position of the foot at the time of impact determine whether the angle becomes more acute or obtuse. The CAG can be disrupted in both directions. BA and the CAG can also be negative in severe injuries.

Orthopedic surgeons frequently use BA and the CAG in reconstructing the normal anatomy of the calcaneus after a fracture. However, the use of the angles in making a diagnosis of a fracture has not been reported in the literature. Chen et al measured BA in 120 patients with normal lateral foot/ankle x-rays in an ED setting and found a range of 148 to 508 with a mean (SD) of 30 F 6 [12]. There was not a significant difference in BA between men and women, left or right feet, and the angle of the central beam. Patients with fractures were not studied.

Gupta et al measured BA and CAG in 25 patients with 27 intraarticular calcaneus fractures and found that the mean (SD) preoperative BA was 15 (14). The mean (SD) CAG in those same patients was 127 (19) [4]. The sensitivity and specificity of the angles in diagnosing calcaneus fractures was not reported in the study. Sensitivity, specificity, and interrelater reliability has not been reported in the literature, to our knowledge.

Our study evaluated the use of BA and the CAG in

65 patients with CT-verified fractures and 68 patients without calcaneus fractures. BA had good interrelater reliability, indicating that the measurement was reproducible among the physicians who participated in the study. Sensitivity and the AUC are most important when using a test to determine if there is a fracture. The AUC for BA was less than 0.9, which leads one to conclude that the sensitivity of using this test alone is inadequate. The CAG measurements were not reproducible, sensitive, or specific. Even when only compression fractures were included in the analysis, performance of the angles did not signifi- cantly improve.

In our study, one fracture was missed on plain films and was only picked up on CT. The fracture was comminuted with significant disruption of bony architecture. It was surprising that the fracture was not visualized on plain films even after we retrospectively reviewed the films and knew a fracture was there.

Patients with calcaneus fractures typically complain of heel pain, inability to bear weight on the affected extremity, swelling of the heel, and ecchymosis tracking toward the sole of the foot. If a high index of clinical suspicion exists for a calcaneus fracture, and plain films are negative, a CT scan should be ordered.

Conclusion

Based on the ROC curves, BA may aid in the diagnosis, and the CAG is not useful in diagnosis of calcaneus fractures in the ED. Emergency physicians were 97.9% accurate in making the diagnosis by reviewing the plain films without bassistanceQ of the angle measurements. Interrater reliability for BA is excellent, but for the CAG, it is poor. Calcaneus fractures may present with normal plain films. A CT scan should be ordered if a physician has a high index of suspicion with negative plain films.

References

1. Harris Jr JH, Harris WH. The radiology of emergency medicine. 2nd ed. Baltimore7 Williams & Wilkins; 1981.
2. Boehler L. Diagnosis, pathology and treatment of fractures of the os calcis. J Bone Joint Surg 1931;13:75 – 89.
3. Sanders R. Current concepts review: displaced intra-articular fractures of the calcaneus. J Bone Joint Surg Am 2000;82-A:225 – 50.
4. Gupta A, Ghalambor N, Nihal A, et al. The modified Palmer lateral approach for calcaneal fractures: wound healing and postoperative computed tomographic evaluation of fracture reduction. Foot Ankle Int 2003;24(10):744 – 53.
5. Schweitzer ME, Karasick D. The foot. In: Rogers LF, editor. Radiology of skeletal trauma. 3rd ed. New York7 Churchill Living- stone, 2002. p. 1332 – 48.
6. Keats TE, Lusted LB. Atlas of roentgenographic measurement. 5th ed. Chicago (IL): Year Book Medical Publishing Inc; 1985.
7. Hauser ML, Kroeker RO. Boehler’s angle: a review and study. J Am Podiatry Assoc 1975;65:517 – 21.
8. Weissman SD. Radiology of the foot. Baltimore7 Williams & Wilkins; 1983.
9. Gissane W. Discussion on bfractures of the os calcisQ. Proceedings of the British Orthopedic Association. J Bone Joint Surg 1947;29:254 – 5.
10. Fitzgibbons TC, McMullen ST, Mattew MA. Fractures and disloca- tions of the calcaneus. In: Bucholz RW, Heckman JD, editors. Rockwood and Green’s fractures in adults, 5th ed. Philadelphia7 Lippincott, Williams and Wilkins, 2001. p. 2134 – 6.
11. Thermann H. Calcaneus fractures: open reduction and internal fixation. In: Wulker N, Stephens M, Cracchiolo III A, editors. An atlas of foot and ankle surgery. St. Louis, Mo7 Mosby, 1998.

p. 233 – 4.

1. Chen M, Bohrer S, Kelley T. Boehler’s angle: a reappraisal. Ann Emerg Med 1991;20(2):122 – 4.