Head Computed Tomography Versus Maxillofacial Computed Tomography An Evaluation of the Efficacy of Facial Imaging in the Detection of Facial Fractures

  In the initial evaluation of a trauma patient, computed tomography (CT) of the head (CTH) is performed to assess for life-threatening intracranial injury. With 25% of all trauma patients sustaining some form of facial injury,¹ many facial fractures are diagnosed, often incidentally, on head CT during the initial evaluation. Whereas maxillofacial CT (CTMF) is widely accepted as the most sensitive method for evaluating facial fractures,² it is often not obtained with the initial trauma survey due to limitations in time and information when assessing a critically ill, severely injured patient. Head CT can be used to diagnose facial fractures in some cases, but failure to obtain dedicated imaging can lead to increased costs related to a missed or delayed facial fracture diagnosis, including increased hospital length of stay.³

Both types of CT imaging are instrumental in evaluating a trauma patient due to their accuracy, reliability, and wide availability. A routine CTH scan is performed in 5-cm axial cuts parallel to the orbitomeatal line from the foramen magnum to the vertex. By comparison, a CTMF scan incorporates 3-cm axial cuts from the mandible to the frontal sinus and is thus able to provide a better look at the facial bones.³ The thinner slices and altered axis gives rise to the diagnostic superiority of CTMF.

Previous studies have identified clinical findings associated with the presence of facial fractures and have subsequently developed clinical criteria to guide CTMF use.^2,4,5 There have also been studies analyzing how head and maxillofacial CT are used in conjunction, recognizing that 84% of facial fracture patients receiving initial trauma CTH required CTMF for definitive diagnosis.³ However, few studies have addressed what specific types of facial fractures are missed on head CT. It is important to note that this study is not insinuating that CTH is an adequate study for the detection of facial fractures.

Because some patients who receive a screening head CT due to trauma are not initially suspected to have a facial fracture or have other critical injuries and do not receive a maxillofacial CT during their initial workup, there are inevitable missed diagnoses. This study seeks to examine what fractures are detected on a head CT (if one is performed prior to CTMF) and what types of fractures would not be detected by reviewing imaging data from patients who presented at a level one trauma center and underwent both head and maxillofacial CT imaging.

A retrospective review of all facial fractures diagnosed and treated at a single institution from 2002 through 2016 was conducted. Inclusion criteria included adults aged 18 years or older, with traumatic etiology, who received CTH and then subsequent CTMF during their initial workup. Patients who had either CTH or CTMF only or a combined CTH/CTMF were excluded. The obtained CTH and CTMF images were examined for facial fractures. Fractures that were missed on CTH and identified on subsequent CTMF were noted. The facial fractures were further subdivided by anatomic location/involvement.

Data organization and basic statistical analysis were conducted using Microsoft Excel (Microsoft Corporation). Statistical significance was set at a level of 5%.

The search criteria yielded 501 patients suffering 1743 total facial fractures who were included in our study. CTH successfully identified 788 (45.21%) fractures, whereas CTMF successfully identified 1743 fractures (100%). In other words, CTH failed to diagnose 955 (54.79%) injuries. Table 1 shows fractures, organized by location, successfully identified on CTH and on CTMF, along with the percentage of successfully identified and missed fractures with respect to that particular fracture location. Furthermore, the Table also outlines the relative diagnostic incidence for a particular fracture location, expressed relative to all fractures identified by the given imaging modality. The most common fractures, in both cohorts, were nasal bone (NB; 273, 15.7%), orbital floor (223, 12.8%), maxillary sinus (214, 12.3%), zygomatic arch (180, 10.3%), medial orbital wall (158, 9.1%), and lateral orbital wall (152, 8.7%). 

Using CTMF to identify missed fractures on CTH, significant differences were noted in the following locations (listed as location, total number of fractures in the given location, number successfully identified on CTH, number missed on CTH and diagnosed on subsequent CTMF, and P value): anterior table frontal sinus (ATFS; 47, 32, 15, P = .002); NB (273, 150, 123, P = .001); pterygoid (53, 16, 37, P = .03); maxillary sinus (214, 127, 87, P < .001); lateral orbital wall (152, 89, 63, P < .001); zygomatic arch (180, 114, 66, P < .001); palate (18, 2, 16, P = 0.004); mandibular symphysis (14, 0, 14, P < .001); mandibular parasymphysis (27, 0, 27, P < .001); mandibular body (67, 5, 62, P < .001); mandibular angle (58, 2, 56, P < .001); mandibular ramus (38, 5, 33, P < .001); mandibular coronoid process (15, 1, 14, P = .003).

No significant differences were noted in fracture diagnosis via CTH versus CTMF for posterior table frontal sinus (PTFS), orbital floor, medial orbital wall, orbital roof, inferior orbital rim, or mandibular condyle.

Figure 1 and Figure 2 show representative images of nasal bone and mandibular fractures initially missed on CTH but diagnosed on subsequent CTMF, respectively.

Prompt identification of facial fractures is crucial because missed or even delayed diagnosis and treatment can result in both functional and aesthetic consequences.⁶ Historically, x-ray was the modality of choice for the initial identification of facial fractures. Compared with CT scans, however, x-rays are less accurate in identifying facial fractures and provide far less detail for potential operative planning.⁶ CT imaging is the current gold standard for the diagnosis of facial fractures, as it is more precise in its characterization of the fractures, aiding in accurate diagnosis and management.^7-10 Despite being the gold standard, CT scans vary in quality due to factors such as manufacturer differences, variations in image sequence or protocol, contrast administration, and patient adherence during the study. Whereas spiral CTs are cost effective, multidetector CTs are faster, with expanded capacity for reconstruction and postprocessing.¹¹ CT scans of the head typically provide only axial cuts and oftentimes are much thicker (about 5 mm) than maxillofacial scans (1 to 3 mm), which can preclude facial fracture identification. Studies have shown that axial CT imaging was significantly more likely to miss a mandibular fracture compared with coronal CT and panoramic radiography.¹⁰

In trauma patients, the priority is to rule out life-threatening injuries that require prompt intervention. Patients typically receive a pan-scan to evaluate for injuries, including CTH, which prevents missed life-threatening intracranial pathology, especially in patients who are unconscious or with reported loss of consciousness.¹² However, this workup does not usually include evaluation of the facial skeleton; therefore, CTH is significantly less likely to identify operative facial fractures compared with facial CT.¹³  A study by Ricci et al concluded that head CT was able to identify the same facial fractures identified by facial CT in only 65% of cases,¹³ compared with only 45.2% in our study. Holmgren et al found that only 16% of patients with facial fractures did not need a facial CT scan after the original head CT scan.³  Because this study excluded patients who obtained only CTH or CTMF, we cannot make any inferences about the proportion of patients for whom CTH alone was sufficient.

Again, to reemphasize, this study does not claim that CTH is adequate for the detection of facial fractures. Rather, because some trauma patients who are not initially suspected to have a facial fracture (or have other critical injuries) get only a screening CTH and no CTMF during their initial workup, there are inevitable missed diagnoses. This study seeks to examine what fractures are detected on a CTH (if performed prior to CTMF) and what types of fractures would not be detected by reviewing imaging data from patients who presented at a level one trauma center and underwent both head and maxillofacial CT imaging.

The use of maxillofacial CT imaging is a vital tool in preoperative planning for surgeons preparing to repair facial fractures. The combined use of coronal, sagittal, and axial imaging of a patient’s facial skeleton allows more precise analysis and operative planning. Moreover, 3-dimensional images that can be produced from CTMF scans are praised as being able to give “an inside out picture of the actual sites of fractures.”⁶ Enhanced preparation prior to operative intervention can lead to decreased instances of unfavorable events and improved patient outcomes. One study claimed that CTMF scans allowed for better surgical treatment in 33% of cases.⁶ Coronal and axial sections were found to be significantly more diagnostic in areas such as orbital floor, arch, lateral maxillary wall, and anterior maxillary wall.⁷ If CTH were the sole imaging modality used in an operative plan for facial fracture reduction, it is likely that a major fracture could be missed due to the lack of coronal/sagittal cuts because CTH alone is prone to missing multiple types of fractures. A 2018 study by Ricci et al analyzed 307 patients receiving both CTH and CTMF scans who underwent subsequent operative facial repair. When compared with CTMF, standard CTH scans failed to identify a significant number of operative facial fractures in 35% of patients. As stated above, CTH missed approximately 955 (54.79%) fractures identified by CTMF in our study.

In our study, CTH was shown to be insufficient for the successful detection of nasal bone fractures. As shown in Table 1, CTH correctly identified only 150 of the 273 (55%) nasal bone fractures. The diagnostic accuracy of the additional 123 (45%) fractures initially missed on CTH and subsequently diagnosed on CTMF was significantly different. Holmgren et al identified nasal fractures as the most common facial fracture discovered by CTH, consistent with this study’s results (about 19% of all fractures identified on CTH, with a true incidence of about 15% identified on CTMF).³

This study also showed CTH was sufficient in diagnosing frontal sinus fractures in some cases, particularly PTFS fractures. Olson et al found that initial CTH identified frontal sinus fractures in 94% of patients.¹⁴ Overall, CTH accurately diagnosed only 54 of 87 (62.07%) of all frontal sinus fractures in our study. When further subdivided by anatomic location into its anterior and posterior table constituents, CTH correctly identified 22 of 40 (55%) PTFS fractures, and the additional CTMF modality did not yield a significantly different diagnostic success; however, CTH was not sufficient in detecting anterior table fractures, missing 15 of the 47 (31.9%) fractures suffered in these patients. Here, CTMF provided a significantly different diagnostic success rate. It is important to note here that if one simply groups all frontal sinus fractures together, one may falsely conclude that CTH is sufficient to identify and diagnose these injuries; however, further breakdown by anatomic location reveals that CTMF may be necessary to accurately and persistently identify fractures.

Interestingly, for bony orbit fractures, CTMF proved significantly useful in diagnosing only lateral orbital wall fractures (CTH missed 41.4% of fractures), but CTH alone was sufficient for orbital floor, medial wall, orbital roof, and inferior orbital rim fractures (CTH success rate = 39.01%, 43.67%, 43.59%, and 43.75%, respectively).

For mandibular fractures, CTH was sufficient only for diagnosing condylar fractures, adequately identifying 12 of 40 (30%) of fractures. CTMF aided in diagnosing the 70% of missed condylar fractures, and while the imaging modality–specific success rates were not significantly different, it is important to note that this difference does approach statistical significance. For all other mandibular fractures (symphysis, parasymphysis, body, angle, ramus, and coronoid), CTH alone was insufficient in yielding diagnostic accuracy and CTMF resulted in a statistically significant difference in fracture identification success rates.

Pterygoid fractures were only successfully diagnosed via CTMF, as only 16 of the 53 (30.12%) fractures were originally identified on CTH. CTMF was significantly more accurate in identifying these fractures. Similar trends were observed for zygomatic arch and palatal fractures.

Of note, it is important to address a linguistic inference that could lead to misinterpretation of these results. What is identified here as a diagnostic success, for example "CTH success rate," is really the instances in which CTH had successfully identified a fracture, meaning that the additional CTMF detection rate was not significantly different in the detection of said fracture. The reason for a seemingly low-percentage success rate could be due to statistics and a low sample size. Furthermore, if the injury were severe enough that it was obvious on CTH, or if the opposite were true in that the injury were so small that even CTMF could not detect it, then the difference in diagnostic accuracy between CTH and CTMF would not be all that different.

One of the issues plaguing the diagnosis of facial fractures is the decision-making process involved when deciding whether a patient should be sent for more extensive imaging after initial vital scans are acquired. Often there is a difference of opinion between clinicians deciding whether a patient, after receiving a CTH, should also be given a CTMF as well. There remains a need to create a decision instrument based on clinical criteria that can ensure appropriate screening in patients who may be at high risk for a facial fracture. Sitzman, Hanson, Alshiek et al from the University of Wisconsin School of Medicine and Public Health created what is now known as the Wisconsin Criteria for screening patients with risk of facial fracture.² The presence of any of 5 physical examination criteria identified a patient as high risk for facial fracture (bony step off or instability, periorbital swelling/contusion, Glasgow Coma Scale less than 14, malocclusion, or tooth absence). The criteria identified all but 6 of the 332 patients with facial fractures in this study, with a reported 98.2% sensitivity for diagnosing facial fractures. However, the Wisconsin Criteria failed in an external study at Yale School of Medicine to replicate a similar sensitivity, instead achieving only 81% sensitivity in all patients who had CT of facial bones over a 6-month period.¹⁵

More work is needed to develop concrete decision-making criteria for high-risk facial fracture patients. For example, etiological aspects of facial fractures could be used to identify high-risk patients. A landmark 11-year study by Hwang et al on facial fractures in 2094 patients from various traumatic etiologies found the most common age group of facial fractures was the third decade (29%), a male predominance of 3.98:1, and the most common etiology was violent assault or nonviolent traumatic injury (49.4%).¹⁶ Inclusion of both physical exam signs and etiological trends into a new criteria for high-risk patients could provide much-needed clarity in future disagreements on imaging strategy.

In recent years, there have been significant advancements in assessing this issue. In 2015, authors investigated clinical predictors that could warrant CTMF in trauma patients who underwent whole-body CT scanning and found that factors such as Glasgow Coma Scale below 8, injury severity scale above 15, intubation, and blood alcohol concentration would indicate abnormal facial findings.¹⁷ The 2017 analysis by Huang et al of concomitant CTH and CTMF in traumatic brain injury patients suggested that traumatic brain injury in trauma patients can be a useful risk factor in necessitating CTMF because routine CTH failed to identify facial fractures that required operative intervention.¹⁸ Other studies suggest that CTH has high diagnostic value, although possibly location dependent; Kim et al’s 2018 investigation suggested that fractures superior to the inferior orbital rim can be accurately diagnosed via CTH.¹⁹ A 2019 investigation out of New Zealand retrospectively analyzed the use of imaging in traumatic facial injuries. The study found that most patients who received plain films also received subsequent CT imaging and that a majority of the patients were managed surgically, indicating that perhaps the patients might have benefitted from receiving CT imaging during their initial workup.²⁰ Similarly in 2020, a group of English researchers proposed an algorithm to identify facial fractures on CTH from trauma patients, with the presence of such factors suggesting the indication to incorporate an additional CTMF during the trauma workup.²¹ Interestingly, an opposite trend has been elucidated in pediatric patient populations. A 2020 study showed that CTH predicted facial injuries with high accuracy in pediatric trauma, and notes that physicians should be prudent to consider facial injuries when reviewing the CTH for initial trauma evaluations.²²

This study demonstrated that for most facial fracture cases, CTMF imaging was optimal in the identification and diagnosis of these injuries because CTH missed many fractures and therefore did not result in operative planning although it was indicated. In the setting of recent studies and findings, this study adds further data and support of the importance of developing standardized protocols to reduce the incidence of missed facial fractures requiring operative intervention.

This study has some important limitations. This analysis examines only 501 patients and thus may be limited in its generalizability and ability to draw inference or conclusions. Furthermore, all healthcare professionals in this study have trained at the same institution. Therefore, their methodology of facial fracture workup and assessment may be similar due to the fact that they were all trained the same way. This introduces reduced variability in imaging reads/protocols, which can influence accurate/successful fracture identification. In other words, because all professionals are at the same institution, their clinical decision-tree may be identical, whereas physicians or residents at other institutions may have different approaches, second to their respective training influences. Additionally, the time interval from initial CTH to subsequent CTMF was not uniform in all patients. All study subjects did receive CTMF on the same admission; however, there was no standardized timepoint between the 2 scans. Likely, the CTMF was performed at a later time due to the severity and critical nature of the patient’s presenting condition. It is not the goal of this article to criticize the fact that the study was not done initially.

Facial fractures can be difficult to diagnose, especially in traumatic injuries in which initial workup does not involve adequate imaging required to diagnose these fractures. Untreated facial fractures can have devastating consequences with lifelong implications, such as impaired functional outcomes and cosmesis. This study showed that overall CTH successfully identified only 45.21% of all injuries sustained in our patients and therefore 54.79% of facial fractures were diagnosed on CTMF scans after CTH readings missed the fractures. CTH alone was only sufficient in detecting PTFS fractures, orbital fractures except of the lateral wall, and mandibular condyle fractures. Future studies should aim to design and validate protocols for facial imaging in trauma patients with suspected facial fractures.

Affiliations: ¹Rutgers New Jersey Medical School, Newark, NJ; ²University of Mississippi Medical Center, Jackson, MS

Correspondence: Zachary Gala; zg60@njms.rutgers.edu

Ethics: This study was performed following IRB approval by the institution; ethical adherence was maintained throughout the duration of the study.

Disclosures: The authors have no relevant financial or nonfinancial interests to disclose.