Eur Arch Otorhinolaryngol DOI 10.1007/s00405-015-3595-8
RHINOLOGY
Ultrasonography in the diagnosis of nasal bone fractures: a comparison with conventional radiography and computed tomography In Sook Lee1 • Jung-Hoon Lee2 • Chang-Ki Woo3 • Hak Jin Kim1 Yu Li Sol1 • Jong Woon Song4 • Kyu-Sup Cho2
•
Received: 6 December 2014 / Accepted: 3 March 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract The purpose of this study was to evaluate and compare the diagnostic efficacy of ultrasonography (US) with radiography and multi-detector computed tomography (CT) for the detection of nasal bone fractures. Forty-one patients with a nasal bone fracture who underwent prospective US examinations were included. Plain radiographs and CT images were obtained on the day of trauma. For US examinations, radiologist used a linear array transducer (L17-5 MHz) in 24 patients and hockey-stick probe (L15-7 MHz) in 17. The bony component of the nose was divided into three parts (right and left lateral nasal walls, and midline of nasal bone). Fracture detection by three modalities was subjected to analysis. Furthermore, findings made by each modality were compared with intraoperative findings. Nasal bone fractures were located in the right lateral wall (n = 28), midline of nasal bone (n = 31), or left lateral wall (n = 31). For right and left lateral nasal walls, CT had greater sensitivity and specificity than US or radiography, and better agreed with intraoperative findings. However, for midline fractures of
& Kyu-Sup Cho [email protected] 1
Department of Radiology and Biomedical Research Institute, Pusan National University School of Medicine, Pusan National University Hospital, Busan, Republic of Korea
2
Department of Otorhinolaryngology and Biomedical Research Institute, Pusan National University School of Medicine, Pusan National University Hospital, 179 GudeokRo, Seo-gu, Busan 602-739, Republic of Korea
3
Department of Otorhinolaryngology, Pusan National University Yangsan Hospital, Yangsan, Republic of Korea
4
Department of Radiology, Inje University Haeundae Paik Hospital, Busan, Republic of Korea
nasal bone, US had higher specificity, positive predictive value, and negative predictive value than CT. Although two US evaluations showed good agreements at all three sites, US findings obtained by the hockey-stick probe showed closer agreement with intraoperative findings for both lateral nasal wall and midline of nasal bone. Although CT showed higher sensitivity and specificity than US or radiography, US found to be helpful for evaluating the midline of nasal bone. Furthermore, for US examinations of the nasal bone, a smaller probe and higher frequency may be required. Keywords Nasal bone Bone fractures Ultrasonography Radiography Multi-detector computed tomography
Introduction Nasal bone fracture is the most commonly encountered facial fracture [1, 2]. Early and accurate recognition and treatment of nasal trauma are important, because incorrect diagnosis and treatment can lead to later deformity [3]. Although clinical examinations are considered standard for diagnosing nasal fractures, hematoma and edema of adjacent tissues make it difficult to diagnose them [4]. Conventional radiography remains the standard imaging procedure, but water’s view targeting the lateral nasal walls is prone to misinterpretation [5, 6]. Computed tomography (CT) is considered a gold standard for diagnosing complex facial fractures, especially mid-facial fractures [7–9]. However, CT is expensive, not readily available, and exposes the patient to high doses of penetrating radiation [4]. Furthermore, CT images are not always good enough to diagnose nasal fractures.
123
Eur Arch Otorhinolaryngol
The value of ultrasonography (US) as a diagnostic tool for detecting fractures has been demonstrated [10–13]. Recently, US has emerged as a useful diagnostic tool for acute nasal fractures, and has been described as an diagnostic alternative, with the advantages of being commonly available, easily used, and free of radiation exposure risks [13]. Higher frequency probes provide higher resolution of the nasal pyramid, especially of the nasal dorsum [14]. The diagnostic merits of high-resolution ultrasonography (HRUS) for the diagnosis of nasal fractures compared with conventional radiography have been demonstrated [10, 13]. The purpose of this study was to evaluate and compare the diagnostic efficacy of US versus conventional radiography and multi-detector CT (MDCT) for the detection of nasal bone fractures based on intraoperative findings. Furthermore, the diagnostic performances of different US probe types were compared.
Materials and methods Subjects Forty-one patients (7 women, 34 men, mean age 29 years, age range 13–66 years) with a nasal bone fracture prospectively underwent US examinations of nasal bones from July 2012 to January 2013. Plain radiographs and CT images were obtained in all cases on the day of trauma. US examinations were performed 1 day before surgery. Average time between injury and US examinations was 12.5 days (10–16 days). Exclusion criteria included a previous history of nasal bone fracture, other maxillofacial fracture that would make a comparison of nasal bone fracture impossible, pregnancy, and an unstable patient condition. This study was approved by the Institutional Review Board of Pusan National University Hospital and informed consent was obtained from all patients. Methods The plain radiography of the nose was performed in lateral and water’s views. CT scans were obtained using multidetector row CT units (SMS Definition or SMS Definition AS?, Siemens Medical Solutions, Erlangen, Germany) with 64 or 128 detector rows. Because all patients had a history of facial trauma, three-dimensional (3D) facial CT was performed. The imaging parameters used were as follows: 120 kVp, 200 mAs and slice thickness 1 mm. Coronal and axial reformatted multiplanar images were obtained from transverse images at a section thickness of 2 mm. Scan times ranged from 10 to 18 s. Two
123
radiologists (K. H. J. and Y. L. S) with 20 and 5 years’ experience of head and neck imaging independently reviewed plain radiographs and MDCT scans, respectively, without knowledge of surgical results and results of each modality. Another one radiologist (L. I. S.) with 10 years’ experience of musculoskeletal US performed the US and analyzed US images without knowledge of the findings of other imaging modalities and clinical data. Two different types of probes were used for the ultrasonographic examination. Twenty-four patients were examined using a L17-5 MHz broadband linear array transducer (iU22, Philips Healthcare) (Fig. 1a) and 17 were examined using a hockey-stick probe (L15-7io broadband compact linear array) (Fig. 1b). In each case, the transducer was placed parallel to the nasal bone directly to skin with applying adequate gel. Ultrasonographic imaging included midline longitudinal imaging and oblique longitudinal scans of lateral walls. For lateral nasal wall examinations, the probe was positioned in an oblique manner and moved sideways and turned to scan to detect possible fracture lines. For midline of nasal bone, the probe was positioned in a vertical plane about the bone. The scanning was started from right lateral nasal wall and next midline, and finished in left lateral nasal wall. Nasal bony components were classified as the right and left lateral nasal walls, and midline of nasal bone. During the analysis, we considered the presence or absence of fracture only, that is, we did not consider fracture severity. The displacement and angulation of the nasal bone or an interruption in continuity were diagnosed as nasal fracture. The three nasal bone sites were analyzed separately. The data obtained from the different imaging modalities were compared with intraoperative findings to obtain sensitivity, specificity and predictive value. Positive intraoperative finding was defined as the depressed fragment to be elevated by means of force in the direction opposite the fracturing force. In addition, agreements between surgical and imaging findings were determined using kappa statistics. Weighted kappa values were used to assess intermethod agreements. The results of weighted kappa values were interpreted as follows: 0.00–0.20 slight agreement, 0.21–0.40 fair, 0.41–0.60 moderate, 0.61–0.80 good, and 0.81–1.00 excellent agreement. Statistical analysis Statistical analysis was performed using the SPSS software package version 21.0 (SPSS Inc., Chicago, IL, http://www. spss.com) and R project to determine sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). A p value \0.05 was considered significant.
Eur Arch Otorhinolaryngol Fig. 1 Types of ultrasonography probe and ultrasonography procedure. The ultrasonography was performed using a broadband linear array transducer (a) or a hockey-stick probe (d). For lateral nasal wall examinations (b, e), the probe was positioned in an oblique manner and moved sideways and turned to scan to detect possible fracture lines. For midline of nasal bone (c, f), the probe was positioned in a vertical plane about the bone
Results All patients underwent closed reduction of fractured nasal bones. Nasal fractures by intraoperative findings were as follows: right lateral nasal wall (n = 28), midline of nasal bone (n = 31), and left lateral nasal wall (n = 31). Twenty-nine patients had a fracture in more than two sites and 21 had a fracture in all three sites. The sensitivity, specificity, PPV, and NPV of the three modalities are summarized in Table 1. For right and left lateral nasal walls, CT had higher sensitivity, specificity, PPV, and NPV than US (linear plus hockey-stick probe findings) (Fig. 2). However, for midline of nasal bone, US had higher specificity, PPV, and NPV than CT, although the sensitivity was same between US and CT. Plain radiographs were found to have lower sensitivity, specificity, PPV, and NPV than US or CT at all three sites (Fig. 3). The agreements between the diagnostic findings of the three modalities and intraoperative findings are summarized in
Table 2. CT showed better agreement with operative findings than US or plane radiography. CT showed good agreement at all sites, although US showed good agreement in the midline of nasal bone and moderate agreement in both lateral nasal walls. In the agreements between the findings of the linear or hockey-stick probes and intraoperative findings, the hockeystick probe showed better agreements with intraoperative findings at all sites (Table 3). Hockey-stick probe showed excellent agreement in the midline of nasal bone and moderate agreement in both lateral nasal walls. However, linear probes showed moderate agreement in the midline of nasal bone and fair agreement in both lateral nasal walls.
Discussion The prompt identification and management of nasal fractures are imperative to avoid the potential complications such as nasal obstruction and posttraumatic nasal deformity
123
Eur Arch Otorhinolaryngol Table 1 Sensitivity, specificity, and positive and negative predictive value of the different modalities
Location Right
Midline
Left
Modality
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
US
76.9 (46.2–95.0)
75.0 (55.1–89.3)
52.6 (28.9–75.6)
86.4 (65.1–97.1)
CT
92.3 (64.0–99.8)
89.3 (71.8–97.7)
76.9 (46.2–95.0)
95.7 (77.5–99.9)
Plain
63.6 (30.8–89.1)
60.9 (38.5–80.3)
43.8 (19.8–70.1)
77.8 (52.4–93.6)
US
80.0 (44.4–97.5)
90.3 (74.3–98.0)
72.7 (39.0–94.0)
93.3 (77.6–99.2)
CT
80.0 (44.4–97.5)
83.9 (66.3–94.6)
66.7 (38.4–88.2)
92.6 (75.7–99.1)
Plain
62.5 (24.5–91.5)
57.7 (36.9–76.7)
57.1 (27.8–83.2)
83.3 (58.6–96.4)
US
70.0 (34.8–93.3)
77.4 (58.9–90.4)
50.0 (23.0–77.0)
88.9 (70.8–97.7)
CT
90.0 (55.5–99.8)
80.7 (62.5–92.6)
60.0 (32.3–83.7)
96.2 (80.4–99.9)
Plain
50.0 (15.7–84.3)
56.0 (34.9–75.6)
30.8 (14.3–51.8)
77.8 (52.4–93.6)
CT computed tomography, NPV negative predictive value, Plain plain radiography, PPV positive predictive value, US ultrasonography
Fig. 2 A 14-year-old male patient with nasal bone fractures. a A longitudinal ultrasonography using broadband linear probe shows a focal cortical discontinuity (arrow) of the left lateral nasal wall. No
fractures were demonstrated at other sites (not shown). b Coronal computed tomography demonstrates fractures (arrows) of the dorsal septum and left lateral nasal wall
[10]. Many diagnostic methods, including clinical examination (crepitus, grossly apparent deviation of the nasal bones, swelling over the nasal bridge, and periorbital ecchymosis), conventional radiography, CT, and US, have been used to evaluate nasal trauma [13]. Although a physical examination is regarded as the gold standard for diagnosis, adequate imaging is often required to address legal implications [13, 15]. Furthermore, physical examination cannot determine fracture complexity [10]. There remains disagreement on which types of imaging modality for nasal fractures are useful [2, 6]. For many years, the standard imaging modality and the initial diagnostic step were considered to be conventional radiography. However, unlike injuries on the midline of nasal bone, it is difficult to evaluate lateral nasal wall injuries on conventional radiographs [1, 15, 16]. Furthermore, several studies have demonstrated that conventional radiography cannot be used routinely to evaluate nasal bone fractures [1, 10, 17]. Previous studies showed that the sensitivity of conventional radiography in detecting fracture line of nasal bone is 79 % [18]. In this study, plain radiography had lower sensitivity, specificity, PPV, and NPV than US or CT at all three sites.
Furthermore, plain radiography showed worse agreement with operative findings than US or CT. The value of ultrasonography (US) as a diagnostic tool for detecting fractures has been demonstrated [10–13]. Recently, US has been introduced as an alternative technique for the evaluation of maxillofacial fractures because it is easy and quick to perform, inexpensive, portable, noninvasive, and does not involve the use of radiation [7, 10, 13]. The diagnostic merits of HRUS for the diagnosis of nasal bone fractures compared with conventional radiography have been demonstrated, especially in pregnant women and child [4, 10, 13, 18, 19]. Furthermore, HRUS was reported to be a reliable diagnostic tool for estimating the time of nasal bone fracture [20] and higher frequency probes provided higher resolution of the nasal pyramid, especially of the nasal dorsum [14]. However, US had some limitations in detecting nondisplaced fractures and revealing the extension of peripheral fracture lines toward central depressions [7, 21]. Similar to our results, several studies have shown that the accuracy of US for the diagnosis of nasal fractures is greater than that of conventional radiography [10, 14].
123
Eur Arch Otorhinolaryngol
Fig. 3 A 22-year-old male patient with nasal bone fractures in all three regions of the nasal pyramid. A longitudinal ultrasonography using hockey-stick probe shows cortical fractures in the right (a), left lateral nasal wall (b), and midline of nasal bone (c). The fracture at
Table 2 Agreements between the diagnostic findings of the three modalities and operative findings
the midline of nasal bone was observed in lateral view of plain radiography (d). Axial computed tomography shows focal angulation of the left nasal wall (e) and coronal view demonstrates midline fracture of nasal bone (f)
Findings
Right lateral nasal wall
Midline of nasal bone
Left lateral nasal wall
Op and US
0.412 (0.007)
0.680 (\0.001)
0.418 (0.006)
Op and CT
0.643 (\0.001)
0.717 (\0.001)
0.604 (\0.001)
Op and plain
0.219 (0.180)
0.150 (0.317)
0.186 (0.066)
Data are expressed as the weighted kappa values (p value) CT computed tomography, Op operative findings, Plain plain radiography, US ultrasonography
Table 3 Agreements between the findings of the linear or hockeystick probes and operative findings Location
Linear probe
Hockey-stick probe
Right lateral nasal wall
0.391 (0.054)
0.430 (0.031)
Midline of nasal bone
0.560 (0.005)
0.850 (\0.001)
Left lateral nasal wall
0.393 (0.041)
0.438 (0.032)
Data are expressed as the weighted kappa values (p value)
Moreover, several studies showed that there is no significant difference between US and CT in the diagnosis of nasal bone fractures [3, 4, 10]. The present study showed that the specificity, PPV, and NPV of US for evaluating the midline of nasal bone were higher than those of CT, but its sensitivity was similar to CT. On the other hand, for the lateral nasal wall, CT achieved a higher sensitivity, specificity, PPV, and NPV than CT. The misinterpretation of the normal nasomaxillary suture line as a fracture and
the lack of a standardized definition of a fracture in this region may contribute to such discrepancies. We believe that the use of thinner axial images and the higher resolution afforded by modern MDCT scanners and software packages explains this difference. Complete assessment of the midline of nasal bone is very important for determining the esthetic and functional outcomes of nasal bone fracture [16]. This study showed that US has higher accuracy rates in the midline fracture of nasal bone than CT or plain radiography. However, in the complex facial bone trauma, CT examinations should be performed [7–9]. Although HRUS is the diagnostic technique of choice when fractures are confined to the nasal bone [22], CT may precisely depict anatomic details of nasal bone and soft tissue, and produce images showing the position and orientation of displaced fractures [10]. Accordingly clinicians frequently find it easier to interpret CT scans than US [17]. However, CT is not always sufficient, and fine nasal fracture lines can be missed due to the partial
123
Eur Arch Otorhinolaryngol
volume artifacts or the direction of fracture or faint fracture [10]. Moreover, CT is expensive, not always available, and delivers high levels of radiation, especially to the lens [10]. In the present study, we used two probes with different shapes and sizes, but similar frequencies, and obtained different results. Interestingly, US images obtained using the smaller hockey-stick probe were more accurate and showed better agreement between US and intraoperative findings than those produced using larger linear probe. A broadband linear array transducer (L17-5 MHz) has a slightly higher frequency compared with a hockey-stick probe (L15-7 MHz). However, the frequencies of both probes are not significantly different and both probes are adequate to examine the bony cortex and have similar resolution. The images obtained by hockey-stick probe are focusing on anatomic sites that we examine due to small field-of-view and show magnified effect. Furthermore, the small size of hockey-stick probe is very effective for evaluating smaller or superficial anatomic sites such as nasal bone. This study has several limitations that require consideration. First, the number of patients was relatively small. Second, the US examiner was aware of patient trauma histories because examinations were performed 1 day before operations. Therefore, this may induce reader bias and result in over estimation for nasal bone fracture. To minimize this effect, the US examiner was unaware of other imaging findings. Third, patients complained of pain or discomfort caused by direct probe contact. The projected external shape of the nose is another limitation of US as usual linear probes are too large to scan nasal bones. Accordingly, we used a smaller hockey-stick probe which was a more suitable size for scanning nasal bones, resulting in reduction of patient complaints. In conclusion, CT showed higher sensitivity, specificity, PPV, and NPV for the evaluation of lateral nasal wall fractures than US or plain radiography. However, US was particularly helpful for evaluating midline of nasal bone. Furthermore, smaller and high frequency US probes may be more helpful in US examinations of the nasal bone. Conflict of interest The authors have no funding, financial relationships, or conflicts of interest to disclose.
References 1. Illum P (1991) Legal aspects in nasal fractures. Rhinology 29:263–266 2. Staffel JG (2002) Optimizing treatment of nasal fractures. Laryngoscope 112:1709–1719 3. Park CH, Joung HH, Lee JH, Hong SM (2009) Usefulness of ultrasonography in the treatment of nasal bone fractures. J Trauma 67:1323–1326
123
4. Javadrashid R, Khatoonabad M, Shams N, Esmaeili F, Jabbari Khamnei H (2011) Comparison of ultrasonography with computed tomography in the diagnosis of nasal bone fractures. Dentomaxillofac Radiol 40:486–491 5. Nigam A, Goni A, Benjamin A, Dasgupta AR (1993) The value of radiographs in the management of the fractured nose. Arch Emerg Med 10:293–297 6. Logan M, O’Driscoll K, Masterson J (1994) The utility of nasal bone radiographs in nasal trauma. Clin Radiol 49:192–194 7. Friedrich RE, Heiland M, Bartel-Friedrich S (2003) Potentials of ultrasound in the diagnosis of midfacial fractures. Clin Oral Investig 7:226–229 8. Jank S, Emshoff R, Etzelsdorfer M, Strobl H, Nicasi A, Norer B (2004) Ultrasound versus computed tomography in the imaging of orbital floor fractures. J Oral Maxillofac Surg 62:150–154 9. Nezafati S, Javadrashid R, Rad S, Akrami S (2010) Comparison of ultrasonography with submentovertex films and computed tomography scan in the diagnosis of zygomatic arch fractures. Dentomaxillofac Radiol 39:11–16 10. Hong HS, Cha JG, Paik SH, Park SJ, Park JS, Kim DH et al (2007) High-resolution sonography for nasal fracture in children. AJR Am J Roentgenol 188:W86–W92 11. Hauger O, Bonnefoy O, Moinard M, Bersani D, Diard F (2002) Occult fractures of the waist of the scaphoid: early diagnosis by high-spatialresolution sonography. AJR Am J Roentgenol 178:1239–1245 12. Paik SH, Chung MJ, Park JS, Goo JM, Im JG (2005) Highresolution sonography of the rib: can fracture and metastasis be differentiated? AJR Am J Roentgenol 184:969–974 13. Thiede O, Kro¨mer JH, Rudack C, Stoll W, Osada N, Schma¨l F (2005) Comparison of ultrasonography and conventional radiography in the diagnosis of nasal fractures. Arch Otolaryngol Head Neck Surg 131:434–439 14. Gu¨rkov R, Clevert D, Krause E (2008) Sonography versus plain x rays in diagnosis of nasal fractures. Am J Rhinol 22:613–616 15. de Lacey GJ, Wignall BK, Hussain S, Reidy JR (1997) The radiology of nasal injuries: problems of interpretation and clinical relevance. Br J Radiol 50:412–414 16. Mondin V, Rinaldo A, Ferlito A (2005) Management of nasal bone fractures. Am J Otolaryngol 26:181–185 17. Daly BD, Russell JL, Davidson MJ, Lamb JT (1990) Thin section computed tomography in the evaluation of naso-ethmoidal trauma. Clin Radiol 41:272–275 18. Mohammadi A, Ghasemi-Rad M (2011) Nasal bone fracture— ultrasonography or computed tomography? Med Ultrason 13:292–295 19. Atighechi S, Baradaranfar MH, Karimi G, Dadgarnia MH, Mansoorian HR, Barkhordari N et al (2014) Diagnostic value of ultrasonography in the diagnosis of nasal fractures. J Craniofac Surg 25:e51–e53 20. Nemati S, Jandaghi AB, Banan R, Aghajanpour M, Kazemnezhad E (2014) Ultrasonography findings in nasal bone fracture; 6-month follow-up: can we estimate time of trauma? Eur Arch Otorhinolaryngol [Epub ahead of print] 21. Hirai T, Manders EK, Nagamoto K, Saggers GC (1996) Ultrasonic observation of facial bone fractures: report of cases. J Oral Maxillofac Surg 54:776–779 22. Lee MH, Cha JG, Hong HS, Lee JS, Park SJ, Paik SH, Lee HK (2009) Comparison of high-resolution ultrasonography and computed tomography in the diagnosis of nasal fractures. J Ultrasound Med 28:717–723
RHINOLOGY
Ultrasonography in the diagnosis of nasal bone fractures: a comparison with conventional radiography and computed tomography In Sook Lee1 • Jung-Hoon Lee2 • Chang-Ki Woo3 • Hak Jin Kim1 Yu Li Sol1 • Jong Woon Song4 • Kyu-Sup Cho2
•
Received: 6 December 2014 / Accepted: 3 March 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract The purpose of this study was to evaluate and compare the diagnostic efficacy of ultrasonography (US) with radiography and multi-detector computed tomography (CT) for the detection of nasal bone fractures. Forty-one patients with a nasal bone fracture who underwent prospective US examinations were included. Plain radiographs and CT images were obtained on the day of trauma. For US examinations, radiologist used a linear array transducer (L17-5 MHz) in 24 patients and hockey-stick probe (L15-7 MHz) in 17. The bony component of the nose was divided into three parts (right and left lateral nasal walls, and midline of nasal bone). Fracture detection by three modalities was subjected to analysis. Furthermore, findings made by each modality were compared with intraoperative findings. Nasal bone fractures were located in the right lateral wall (n = 28), midline of nasal bone (n = 31), or left lateral wall (n = 31). For right and left lateral nasal walls, CT had greater sensitivity and specificity than US or radiography, and better agreed with intraoperative findings. However, for midline fractures of
& Kyu-Sup Cho [email protected] 1
Department of Radiology and Biomedical Research Institute, Pusan National University School of Medicine, Pusan National University Hospital, Busan, Republic of Korea
2
Department of Otorhinolaryngology and Biomedical Research Institute, Pusan National University School of Medicine, Pusan National University Hospital, 179 GudeokRo, Seo-gu, Busan 602-739, Republic of Korea
3
Department of Otorhinolaryngology, Pusan National University Yangsan Hospital, Yangsan, Republic of Korea
4
Department of Radiology, Inje University Haeundae Paik Hospital, Busan, Republic of Korea
nasal bone, US had higher specificity, positive predictive value, and negative predictive value than CT. Although two US evaluations showed good agreements at all three sites, US findings obtained by the hockey-stick probe showed closer agreement with intraoperative findings for both lateral nasal wall and midline of nasal bone. Although CT showed higher sensitivity and specificity than US or radiography, US found to be helpful for evaluating the midline of nasal bone. Furthermore, for US examinations of the nasal bone, a smaller probe and higher frequency may be required. Keywords Nasal bone Bone fractures Ultrasonography Radiography Multi-detector computed tomography
Introduction Nasal bone fracture is the most commonly encountered facial fracture [1, 2]. Early and accurate recognition and treatment of nasal trauma are important, because incorrect diagnosis and treatment can lead to later deformity [3]. Although clinical examinations are considered standard for diagnosing nasal fractures, hematoma and edema of adjacent tissues make it difficult to diagnose them [4]. Conventional radiography remains the standard imaging procedure, but water’s view targeting the lateral nasal walls is prone to misinterpretation [5, 6]. Computed tomography (CT) is considered a gold standard for diagnosing complex facial fractures, especially mid-facial fractures [7–9]. However, CT is expensive, not readily available, and exposes the patient to high doses of penetrating radiation [4]. Furthermore, CT images are not always good enough to diagnose nasal fractures.
123
Eur Arch Otorhinolaryngol
The value of ultrasonography (US) as a diagnostic tool for detecting fractures has been demonstrated [10–13]. Recently, US has emerged as a useful diagnostic tool for acute nasal fractures, and has been described as an diagnostic alternative, with the advantages of being commonly available, easily used, and free of radiation exposure risks [13]. Higher frequency probes provide higher resolution of the nasal pyramid, especially of the nasal dorsum [14]. The diagnostic merits of high-resolution ultrasonography (HRUS) for the diagnosis of nasal fractures compared with conventional radiography have been demonstrated [10, 13]. The purpose of this study was to evaluate and compare the diagnostic efficacy of US versus conventional radiography and multi-detector CT (MDCT) for the detection of nasal bone fractures based on intraoperative findings. Furthermore, the diagnostic performances of different US probe types were compared.
Materials and methods Subjects Forty-one patients (7 women, 34 men, mean age 29 years, age range 13–66 years) with a nasal bone fracture prospectively underwent US examinations of nasal bones from July 2012 to January 2013. Plain radiographs and CT images were obtained in all cases on the day of trauma. US examinations were performed 1 day before surgery. Average time between injury and US examinations was 12.5 days (10–16 days). Exclusion criteria included a previous history of nasal bone fracture, other maxillofacial fracture that would make a comparison of nasal bone fracture impossible, pregnancy, and an unstable patient condition. This study was approved by the Institutional Review Board of Pusan National University Hospital and informed consent was obtained from all patients. Methods The plain radiography of the nose was performed in lateral and water’s views. CT scans were obtained using multidetector row CT units (SMS Definition or SMS Definition AS?, Siemens Medical Solutions, Erlangen, Germany) with 64 or 128 detector rows. Because all patients had a history of facial trauma, three-dimensional (3D) facial CT was performed. The imaging parameters used were as follows: 120 kVp, 200 mAs and slice thickness 1 mm. Coronal and axial reformatted multiplanar images were obtained from transverse images at a section thickness of 2 mm. Scan times ranged from 10 to 18 s. Two
123
radiologists (K. H. J. and Y. L. S) with 20 and 5 years’ experience of head and neck imaging independently reviewed plain radiographs and MDCT scans, respectively, without knowledge of surgical results and results of each modality. Another one radiologist (L. I. S.) with 10 years’ experience of musculoskeletal US performed the US and analyzed US images without knowledge of the findings of other imaging modalities and clinical data. Two different types of probes were used for the ultrasonographic examination. Twenty-four patients were examined using a L17-5 MHz broadband linear array transducer (iU22, Philips Healthcare) (Fig. 1a) and 17 were examined using a hockey-stick probe (L15-7io broadband compact linear array) (Fig. 1b). In each case, the transducer was placed parallel to the nasal bone directly to skin with applying adequate gel. Ultrasonographic imaging included midline longitudinal imaging and oblique longitudinal scans of lateral walls. For lateral nasal wall examinations, the probe was positioned in an oblique manner and moved sideways and turned to scan to detect possible fracture lines. For midline of nasal bone, the probe was positioned in a vertical plane about the bone. The scanning was started from right lateral nasal wall and next midline, and finished in left lateral nasal wall. Nasal bony components were classified as the right and left lateral nasal walls, and midline of nasal bone. During the analysis, we considered the presence or absence of fracture only, that is, we did not consider fracture severity. The displacement and angulation of the nasal bone or an interruption in continuity were diagnosed as nasal fracture. The three nasal bone sites were analyzed separately. The data obtained from the different imaging modalities were compared with intraoperative findings to obtain sensitivity, specificity and predictive value. Positive intraoperative finding was defined as the depressed fragment to be elevated by means of force in the direction opposite the fracturing force. In addition, agreements between surgical and imaging findings were determined using kappa statistics. Weighted kappa values were used to assess intermethod agreements. The results of weighted kappa values were interpreted as follows: 0.00–0.20 slight agreement, 0.21–0.40 fair, 0.41–0.60 moderate, 0.61–0.80 good, and 0.81–1.00 excellent agreement. Statistical analysis Statistical analysis was performed using the SPSS software package version 21.0 (SPSS Inc., Chicago, IL, http://www. spss.com) and R project to determine sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). A p value \0.05 was considered significant.
Eur Arch Otorhinolaryngol Fig. 1 Types of ultrasonography probe and ultrasonography procedure. The ultrasonography was performed using a broadband linear array transducer (a) or a hockey-stick probe (d). For lateral nasal wall examinations (b, e), the probe was positioned in an oblique manner and moved sideways and turned to scan to detect possible fracture lines. For midline of nasal bone (c, f), the probe was positioned in a vertical plane about the bone
Results All patients underwent closed reduction of fractured nasal bones. Nasal fractures by intraoperative findings were as follows: right lateral nasal wall (n = 28), midline of nasal bone (n = 31), and left lateral nasal wall (n = 31). Twenty-nine patients had a fracture in more than two sites and 21 had a fracture in all three sites. The sensitivity, specificity, PPV, and NPV of the three modalities are summarized in Table 1. For right and left lateral nasal walls, CT had higher sensitivity, specificity, PPV, and NPV than US (linear plus hockey-stick probe findings) (Fig. 2). However, for midline of nasal bone, US had higher specificity, PPV, and NPV than CT, although the sensitivity was same between US and CT. Plain radiographs were found to have lower sensitivity, specificity, PPV, and NPV than US or CT at all three sites (Fig. 3). The agreements between the diagnostic findings of the three modalities and intraoperative findings are summarized in
Table 2. CT showed better agreement with operative findings than US or plane radiography. CT showed good agreement at all sites, although US showed good agreement in the midline of nasal bone and moderate agreement in both lateral nasal walls. In the agreements between the findings of the linear or hockey-stick probes and intraoperative findings, the hockeystick probe showed better agreements with intraoperative findings at all sites (Table 3). Hockey-stick probe showed excellent agreement in the midline of nasal bone and moderate agreement in both lateral nasal walls. However, linear probes showed moderate agreement in the midline of nasal bone and fair agreement in both lateral nasal walls.
Discussion The prompt identification and management of nasal fractures are imperative to avoid the potential complications such as nasal obstruction and posttraumatic nasal deformity
123
Eur Arch Otorhinolaryngol Table 1 Sensitivity, specificity, and positive and negative predictive value of the different modalities
Location Right
Midline
Left
Modality
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
US
76.9 (46.2–95.0)
75.0 (55.1–89.3)
52.6 (28.9–75.6)
86.4 (65.1–97.1)
CT
92.3 (64.0–99.8)
89.3 (71.8–97.7)
76.9 (46.2–95.0)
95.7 (77.5–99.9)
Plain
63.6 (30.8–89.1)
60.9 (38.5–80.3)
43.8 (19.8–70.1)
77.8 (52.4–93.6)
US
80.0 (44.4–97.5)
90.3 (74.3–98.0)
72.7 (39.0–94.0)
93.3 (77.6–99.2)
CT
80.0 (44.4–97.5)
83.9 (66.3–94.6)
66.7 (38.4–88.2)
92.6 (75.7–99.1)
Plain
62.5 (24.5–91.5)
57.7 (36.9–76.7)
57.1 (27.8–83.2)
83.3 (58.6–96.4)
US
70.0 (34.8–93.3)
77.4 (58.9–90.4)
50.0 (23.0–77.0)
88.9 (70.8–97.7)
CT
90.0 (55.5–99.8)
80.7 (62.5–92.6)
60.0 (32.3–83.7)
96.2 (80.4–99.9)
Plain
50.0 (15.7–84.3)
56.0 (34.9–75.6)
30.8 (14.3–51.8)
77.8 (52.4–93.6)
CT computed tomography, NPV negative predictive value, Plain plain radiography, PPV positive predictive value, US ultrasonography
Fig. 2 A 14-year-old male patient with nasal bone fractures. a A longitudinal ultrasonography using broadband linear probe shows a focal cortical discontinuity (arrow) of the left lateral nasal wall. No
fractures were demonstrated at other sites (not shown). b Coronal computed tomography demonstrates fractures (arrows) of the dorsal septum and left lateral nasal wall
[10]. Many diagnostic methods, including clinical examination (crepitus, grossly apparent deviation of the nasal bones, swelling over the nasal bridge, and periorbital ecchymosis), conventional radiography, CT, and US, have been used to evaluate nasal trauma [13]. Although a physical examination is regarded as the gold standard for diagnosis, adequate imaging is often required to address legal implications [13, 15]. Furthermore, physical examination cannot determine fracture complexity [10]. There remains disagreement on which types of imaging modality for nasal fractures are useful [2, 6]. For many years, the standard imaging modality and the initial diagnostic step were considered to be conventional radiography. However, unlike injuries on the midline of nasal bone, it is difficult to evaluate lateral nasal wall injuries on conventional radiographs [1, 15, 16]. Furthermore, several studies have demonstrated that conventional radiography cannot be used routinely to evaluate nasal bone fractures [1, 10, 17]. Previous studies showed that the sensitivity of conventional radiography in detecting fracture line of nasal bone is 79 % [18]. In this study, plain radiography had lower sensitivity, specificity, PPV, and NPV than US or CT at all three sites.
Furthermore, plain radiography showed worse agreement with operative findings than US or CT. The value of ultrasonography (US) as a diagnostic tool for detecting fractures has been demonstrated [10–13]. Recently, US has been introduced as an alternative technique for the evaluation of maxillofacial fractures because it is easy and quick to perform, inexpensive, portable, noninvasive, and does not involve the use of radiation [7, 10, 13]. The diagnostic merits of HRUS for the diagnosis of nasal bone fractures compared with conventional radiography have been demonstrated, especially in pregnant women and child [4, 10, 13, 18, 19]. Furthermore, HRUS was reported to be a reliable diagnostic tool for estimating the time of nasal bone fracture [20] and higher frequency probes provided higher resolution of the nasal pyramid, especially of the nasal dorsum [14]. However, US had some limitations in detecting nondisplaced fractures and revealing the extension of peripheral fracture lines toward central depressions [7, 21]. Similar to our results, several studies have shown that the accuracy of US for the diagnosis of nasal fractures is greater than that of conventional radiography [10, 14].
123
Eur Arch Otorhinolaryngol
Fig. 3 A 22-year-old male patient with nasal bone fractures in all three regions of the nasal pyramid. A longitudinal ultrasonography using hockey-stick probe shows cortical fractures in the right (a), left lateral nasal wall (b), and midline of nasal bone (c). The fracture at
Table 2 Agreements between the diagnostic findings of the three modalities and operative findings
the midline of nasal bone was observed in lateral view of plain radiography (d). Axial computed tomography shows focal angulation of the left nasal wall (e) and coronal view demonstrates midline fracture of nasal bone (f)
Findings
Right lateral nasal wall
Midline of nasal bone
Left lateral nasal wall
Op and US
0.412 (0.007)
0.680 (\0.001)
0.418 (0.006)
Op and CT
0.643 (\0.001)
0.717 (\0.001)
0.604 (\0.001)
Op and plain
0.219 (0.180)
0.150 (0.317)
0.186 (0.066)
Data are expressed as the weighted kappa values (p value) CT computed tomography, Op operative findings, Plain plain radiography, US ultrasonography
Table 3 Agreements between the findings of the linear or hockeystick probes and operative findings Location
Linear probe
Hockey-stick probe
Right lateral nasal wall
0.391 (0.054)
0.430 (0.031)
Midline of nasal bone
0.560 (0.005)
0.850 (\0.001)
Left lateral nasal wall
0.393 (0.041)
0.438 (0.032)
Data are expressed as the weighted kappa values (p value)
Moreover, several studies showed that there is no significant difference between US and CT in the diagnosis of nasal bone fractures [3, 4, 10]. The present study showed that the specificity, PPV, and NPV of US for evaluating the midline of nasal bone were higher than those of CT, but its sensitivity was similar to CT. On the other hand, for the lateral nasal wall, CT achieved a higher sensitivity, specificity, PPV, and NPV than CT. The misinterpretation of the normal nasomaxillary suture line as a fracture and
the lack of a standardized definition of a fracture in this region may contribute to such discrepancies. We believe that the use of thinner axial images and the higher resolution afforded by modern MDCT scanners and software packages explains this difference. Complete assessment of the midline of nasal bone is very important for determining the esthetic and functional outcomes of nasal bone fracture [16]. This study showed that US has higher accuracy rates in the midline fracture of nasal bone than CT or plain radiography. However, in the complex facial bone trauma, CT examinations should be performed [7–9]. Although HRUS is the diagnostic technique of choice when fractures are confined to the nasal bone [22], CT may precisely depict anatomic details of nasal bone and soft tissue, and produce images showing the position and orientation of displaced fractures [10]. Accordingly clinicians frequently find it easier to interpret CT scans than US [17]. However, CT is not always sufficient, and fine nasal fracture lines can be missed due to the partial
123
Eur Arch Otorhinolaryngol
volume artifacts or the direction of fracture or faint fracture [10]. Moreover, CT is expensive, not always available, and delivers high levels of radiation, especially to the lens [10]. In the present study, we used two probes with different shapes and sizes, but similar frequencies, and obtained different results. Interestingly, US images obtained using the smaller hockey-stick probe were more accurate and showed better agreement between US and intraoperative findings than those produced using larger linear probe. A broadband linear array transducer (L17-5 MHz) has a slightly higher frequency compared with a hockey-stick probe (L15-7 MHz). However, the frequencies of both probes are not significantly different and both probes are adequate to examine the bony cortex and have similar resolution. The images obtained by hockey-stick probe are focusing on anatomic sites that we examine due to small field-of-view and show magnified effect. Furthermore, the small size of hockey-stick probe is very effective for evaluating smaller or superficial anatomic sites such as nasal bone. This study has several limitations that require consideration. First, the number of patients was relatively small. Second, the US examiner was aware of patient trauma histories because examinations were performed 1 day before operations. Therefore, this may induce reader bias and result in over estimation for nasal bone fracture. To minimize this effect, the US examiner was unaware of other imaging findings. Third, patients complained of pain or discomfort caused by direct probe contact. The projected external shape of the nose is another limitation of US as usual linear probes are too large to scan nasal bones. Accordingly, we used a smaller hockey-stick probe which was a more suitable size for scanning nasal bones, resulting in reduction of patient complaints. In conclusion, CT showed higher sensitivity, specificity, PPV, and NPV for the evaluation of lateral nasal wall fractures than US or plain radiography. However, US was particularly helpful for evaluating midline of nasal bone. Furthermore, smaller and high frequency US probes may be more helpful in US examinations of the nasal bone. Conflict of interest The authors have no funding, financial relationships, or conflicts of interest to disclose.
References 1. Illum P (1991) Legal aspects in nasal fractures. Rhinology 29:263–266 2. Staffel JG (2002) Optimizing treatment of nasal fractures. Laryngoscope 112:1709–1719 3. Park CH, Joung HH, Lee JH, Hong SM (2009) Usefulness of ultrasonography in the treatment of nasal bone fractures. J Trauma 67:1323–1326
123
4. Javadrashid R, Khatoonabad M, Shams N, Esmaeili F, Jabbari Khamnei H (2011) Comparison of ultrasonography with computed tomography in the diagnosis of nasal bone fractures. Dentomaxillofac Radiol 40:486–491 5. Nigam A, Goni A, Benjamin A, Dasgupta AR (1993) The value of radiographs in the management of the fractured nose. Arch Emerg Med 10:293–297 6. Logan M, O’Driscoll K, Masterson J (1994) The utility of nasal bone radiographs in nasal trauma. Clin Radiol 49:192–194 7. Friedrich RE, Heiland M, Bartel-Friedrich S (2003) Potentials of ultrasound in the diagnosis of midfacial fractures. Clin Oral Investig 7:226–229 8. Jank S, Emshoff R, Etzelsdorfer M, Strobl H, Nicasi A, Norer B (2004) Ultrasound versus computed tomography in the imaging of orbital floor fractures. J Oral Maxillofac Surg 62:150–154 9. Nezafati S, Javadrashid R, Rad S, Akrami S (2010) Comparison of ultrasonography with submentovertex films and computed tomography scan in the diagnosis of zygomatic arch fractures. Dentomaxillofac Radiol 39:11–16 10. Hong HS, Cha JG, Paik SH, Park SJ, Park JS, Kim DH et al (2007) High-resolution sonography for nasal fracture in children. AJR Am J Roentgenol 188:W86–W92 11. Hauger O, Bonnefoy O, Moinard M, Bersani D, Diard F (2002) Occult fractures of the waist of the scaphoid: early diagnosis by high-spatialresolution sonography. AJR Am J Roentgenol 178:1239–1245 12. Paik SH, Chung MJ, Park JS, Goo JM, Im JG (2005) Highresolution sonography of the rib: can fracture and metastasis be differentiated? AJR Am J Roentgenol 184:969–974 13. Thiede O, Kro¨mer JH, Rudack C, Stoll W, Osada N, Schma¨l F (2005) Comparison of ultrasonography and conventional radiography in the diagnosis of nasal fractures. Arch Otolaryngol Head Neck Surg 131:434–439 14. Gu¨rkov R, Clevert D, Krause E (2008) Sonography versus plain x rays in diagnosis of nasal fractures. Am J Rhinol 22:613–616 15. de Lacey GJ, Wignall BK, Hussain S, Reidy JR (1997) The radiology of nasal injuries: problems of interpretation and clinical relevance. Br J Radiol 50:412–414 16. Mondin V, Rinaldo A, Ferlito A (2005) Management of nasal bone fractures. Am J Otolaryngol 26:181–185 17. Daly BD, Russell JL, Davidson MJ, Lamb JT (1990) Thin section computed tomography in the evaluation of naso-ethmoidal trauma. Clin Radiol 41:272–275 18. Mohammadi A, Ghasemi-Rad M (2011) Nasal bone fracture— ultrasonography or computed tomography? Med Ultrason 13:292–295 19. Atighechi S, Baradaranfar MH, Karimi G, Dadgarnia MH, Mansoorian HR, Barkhordari N et al (2014) Diagnostic value of ultrasonography in the diagnosis of nasal fractures. J Craniofac Surg 25:e51–e53 20. Nemati S, Jandaghi AB, Banan R, Aghajanpour M, Kazemnezhad E (2014) Ultrasonography findings in nasal bone fracture; 6-month follow-up: can we estimate time of trauma? Eur Arch Otorhinolaryngol [Epub ahead of print] 21. Hirai T, Manders EK, Nagamoto K, Saggers GC (1996) Ultrasonic observation of facial bone fractures: report of cases. J Oral Maxillofac Surg 54:776–779 22. Lee MH, Cha JG, Hong HS, Lee JS, Park SJ, Paik SH, Lee HK (2009) Comparison of high-resolution ultrasonography and computed tomography in the diagnosis of nasal fractures. J Ultrasound Med 28:717–723
