NOE indicates nasoorbitoethmoid.
Radabaugh JP, Zhang P, Wang D, Lin PLY, Shelton J, Liau JY, Cunningham LL, Gal TJ. Barriers to Repair in Maxillofacial Trauma. JAMA Facial Plast Surg. 2016;18(3):177-182. doi:10.1001/jamafacial.2015.2101
Multiple factors can be associated with the delayed repair of maxillofacial injuries that may be associated with increased morbidity.
To assess factors affecting timing of repair and barriers which may exist in the management of maxillofacial trauma.
Design, Setting, and Participants
This retrospective cohort study at a tertiary care facility used the Current Procedural Terminology coding to identify adult patients undergoing operative repair of maxillofacial injuries between January 2010 and December 2013. Demographic information, presence and severity of concomitant injuries, as well as fracture-specific data including fracture type(s), mechanism of injury, and documented complications were recorded. Identifiable delays for medical, logistical, or other reasons were also documented. Multivariate regression modeling was used to determine factors associated with increased time to repair. A comparative analysis was used to identify association between complications and time to operative repair.
Main Outcomes and Measures
Time to operative repair from date of presentation; association of known operative delay and perioperative complications.
Overall, 780 patients were included in the study. Of patients meeting inclusion criteria, mean (SD) age was 36.7 (14.2) years (range, 18-88 years), and 616 patients (79%) were male. Average time to repair was 6.5 days (range, 0-43 days), and 138 patients (17.7%) were observed to have a documented reason for delay for medical reasons (n = 62 [44.9%]), operating room logistical factors (n = 17 [12.3%]), or other reasons (n = 59 patients [42.8%]) either as a function of delayed patient presentation or failure of patients to make scheduled appointments or operations. Injury severity score (ρ = 0.45; P < .001), concurrent injuries (P < .001), decreased Glasgow Coma Scale (P < .001) and inpatient status at time of surgery (P < .001), were associated with increased time to repair. The observed complication rate was 13.6%. There was no statistically significant association between known operative delay and development of complications (χ21 = 2.92; P = .08).
Conclusions and Relevance
Management of maxillofacial trauma appears to occur in a timely manner. Patient injury severity appears to have the greatest effect on timing of repair. While delays in operative repair may be unavoidable in certain circumstances, streamlining and managing causes of known delay may help improve and expedite patient care.
Level of Evidence
Maxillofacial trauma accounts for over 400 000 emergency department visits annually.1 The goals of fracture management are the return of form and function with minimal complications. Postoperative complication rates in the management of maxillofacial trauma range from 5% to 32%, with significant variations by fracture site.2 There is some evidence to suggest that timely management of fracture repair is beneficial in reducing morbidity.
Optimal timing of maxillofacial trauma repair remains a source of some controversy in the literature. There is a large amount of evidence to support expedient fracture repair as a contributing factor in reducing postoperative complications.2- 6 On the contrary, there is almost an equal amount of evidence to suggest that timing of repair may not make a difference.7- 10 To further complicate the issue, even the definition of an early vs a late operative repair varies in the literature. Frequently, studies have defined this based on different anecdotal criteria, specifics of surgeon experience, or wound healing principles.
Many times the timing of repair is beyond the control of the surgeon. More severe concomitant injuries may delay the opportunity to expediently perform repair. Similarly, logistical concerns such as operating room availability, surgeon availability, as well as patient scheduling and compliance may complicate the ability to address injuries within what is considered to be an appropriate timeframe. While these factors are frequently cited in the literature as potential reasons for delayed repair, there has been limited investigation into the roles these factors truly play in the management of facial trauma.
The primary aim of this study is to identify the barriers that exist in the operative management of maxillofacial trauma. Using a retrospective cohort from a tertiary care level I trauma center, we seek to examine the association of injury specifics, associated injuries, hospital and/or operative resources, patient scheduling and compliance, and the timing of operative repair. In addition, we will examine complication rates as a function of time to repair.
This study was performed after approval by the University of Kentucky institutional review board (IRB). Data was collected retrospectively through review of medical records. Patients included in the study underwent operative repair of maxillofacial fractures at the University of Kentucky Medical Center between January 2009 and December 2013. Cases were identified by querying hospital operative records using Current Procedural Terminology codes for all cases of operative maxillofacial trauma. This included all fractures of the mandible, midface (including LeFort fractures), zygomaticomaxillary complex, zygoma, nasoorbitoethmoid, and frontal sinus were included. All codes used for maxillofacial fracture repair were used. The study included fractures repaired by plastic surgery, otolaryngology, and oral and maxillofacial surgery, among which trauma is equally shared at this institution. Any cases performed by the oculoplastics service were also included. Nasal fractures and isolated soft tissue trauma were excluded. Patients younger than 18 years, pregnant females, preexisting facial pathology, and ballistic trauma were also excluded. Facial fractures intentionally managed nonoperatively were also excluded. Patients with medical records incomplete or insufficient for inclusion were also excluded. As the primary goal of this study was to examine delays in repair, fractures requiring emergent management were also excluded, as it would be presumed that repair would occur on the day of injury or presentation. Indications for emergent repair included orbital entrapment, uncontrolled hemorrhage, or significant intracranial injury involving the fracture site.
Demographic information collected included patient age, sex, and insurance status. Injury-specific variables included mechanism of injury, concomitant injuries, injury severity score,11 abbreviated injury score for the head and neck, Glasgow Coma Scale (GCS), and inpatient status. Mechanism of injury was categorized as motor vehicle crash, motorcycle crash, all-terrain vehicle crash, pedestrian vs vehicle crash, fall, assault, or work-related injury. Concomitant injuries were categorized as neurologic (intracranial), neurologic (spine), orthopedic, abdominal, thoracic, vascular, or other. Maxillofacial fractures were categorized as isolated, multiple, or panfacial. Isolated fractures were categorized as frontal sinus, zygomaticomaxillary (including isolated zygoma fractures), orbital (floor), nasoorbitoethmoid, mandible, and LeFort fractures. LeFort fractures were classified using standard definitions for these fractures, whereas isolated fractures of the nasoorbitoethmoid complex or zygomaticomaxillary complex were categorized separately. Mandible fractures with multiple fracture sites isolated to the mandible were considered isolated mandible fractures. Fractures were considered multiple if they involved any 2 isolated fracture sites but were not considered panfacial fractures. To be classified as a panfacial fracture, fractures of the mandible, midface, and frontal region (frontal sinus fracture or nasoorbitoethmoid complex fracture) had to be present.
Date of injury, date of presentation, and date of operation were also recorded. For the purposes of this analysis, time 0 was considered date of presentation to the health care facility. Time to repair was calculated from date of presentation to date of operation. Cases were also classified as to whether a known delay in surgery existed. Known delays were categorized owing to medical condition, operating room logistics, or other. Other delays included scheduling difficulties, patients not showing for scheduled surgery, or missed follow up appointments following emergency department evaluation.
Statistical analysis was performed using SAS version 9.4 (SAS Institute, Inc). Descriptive statistics were generated for the cohort (Table 1). Multivariate linear regression was used to examine the association with factors associated with delay in repair as a function of time. Multiple logistic regression was also performed to evaluate the association of delayed repair with the development of postoperative complications. Group differences for continuous variables were tested with t tests, and group differences for categorical variables were tested with χ2 tests.
Overall, 874 patients with operative maxillofacial trauma were identified during the study period. After exclusion, 780 met criteria for inclusion in the study. Demographics for the cohort are presented in Table 1. Mean (SD) age was 36.7 (14.2) years (range, 18-88 years), and 616 patients (79%) were male. The distribution of mechanism of injury is illustrated in Figure 1. Assault (n = 322 [41.3%]) and motor vehicle crash (n = 157 [20.5%]) were the predominant cause of injury. Frequency of fracture type is presented in Figure 2. Mandible fractures were by far the most common fracture type, comprising 43.6% (n = 340) of the cohort.
Mean (SD) time to repair for the entire cohort was 6.5 (5.0) days (range, 0-43 days). Mean time to repair by fracture type is also listed in Table 2. Isolated LeFort fractures were found to have the longest interval from injury to repair, with a mean of 23.8 days. Conversely, isolated nasoorbitoethmoid fractures had the shortest interval between presentation and repair with a mean of 5 days. Of the 780 patients included, 273 (35%) were admitted at the time of presentation, and 240 (30.8%) remained admitted at the time of surgery. Mean time to repair for patients classified as inpatient at the time of repair was 4.9 days compared with 7.1 days for outpatients. Of all cases, 362 (46.4%) were scheduled in advance, while 382 patients (49.4%) were placed on the operating room waiting list, classified as “space available” or “first come, first served.” Less than 5% of patients were posted as a classified emergency. This is to be distinguished from cases with specific, acute emergent symptoms that were excluded prior to analysis.
Overall, 138 patients (17.7%) were classified as having a known delay in repair or that there was a delay in the repair of the fracture for specific reasons beyond those the surgical team could control. There was a fairly equivalent distribution between known delays considered to be for medical reasons (n = 62 [44.9%]) and those classified as having other reasons (n = 59 [42.8%]). Of the 59 patients considered to have other reasons for delay, 14 did not show up on the day of a scheduled operation; 21 patients presented to outside facilities and were given close outpatient follow-up for evaluation but failed to make this initial appointment; and 15 patients did not show up to the hospital after being accepted in transfer. These 15 patients who had been accepted but did not show were either identified by the on-call clinician who accepted a patient in transfer from an outside hospital or accepted in hospital to hospital transfer where delays in transfer were encountered secondary to bed availability. Seven patients in this group (almost exclusively isolated subcondylar fractures) failed a trial of conservative therapy. Operating room (OR) logistical factors were considered to be situations where cases could not be scheduled in a timely manner owing to OR availability, cases that were scheduled but could not be completed owing to OR staffing (nursing or anesthesia) insufficiencies, or cases posted and placed on the “space available” list where space was not available for longer than 24 hours. Operating room logistics were identified as the cause for delay in 17 cases (12.3%) with known delay. Mean (SD) time to repair in patients classified as having a known delay was 9.8 (5.9) days compared with 5.7 (4.5) days for the balance of the cohort (P < .001).
Multivariate analysis was performed to account for interactions between variables and their association with timing of repair. Linear regression modeling was performed using the following categorical variables: fracture type, mechanism of injury, presence of concurrent injuries, GCS, admission status at time of presentation compared with time of surgery, and presence of a known delay (Table 3). Isolated mandible fractures (P = .01) as well as orbital floor fractures (P = .02) were observed to have a statistically significant association with delayed repair. Concomitant injuries, decreased GCS, and inpatient status at the time of surgery were all associated with delayed repair (P < .001). It stands to reason that these variables are, to some degree, interrelated. However, all 5 remained statistically significant in univariate modeling as well as multiple iterations of multivariate modeling. Not surprisingly, the presence of a known operative delay was associated statistically with a delayed repair (P < .001). Removing these patients from the model did not affect other findings. No significant association was observed between type of known delay and timing of repair (P = .08).
Complications occurred in 106 patients (13.6%). Nearly half (n = 50 [47.2%]) were postoperative infections. Distribution of complications is listed in Table 4. Mean (SD) time to repair in patients developing complications was 5.4 (4.4) days, vs 6.6 (5.1) days for patients without complications, which was statistically significant (P = .02). A trend toward increased complications was observed in patients with a known operative delay; however, this was not statistically significant (P = .08). There is some evidence to suggest fractures should be repaired within 3 days.3- 6 For this reason, we created a dichotomous variable comparing patients repaired either before or after 2.5 days. For patients treated in less than 2.5 days, the complication rates were 19.6% (n = 33 of 168) compared with 11.9% (n = 73 of 612) for those treated after 2.5 days (P = .02), which suggests no disadvantage to clinically appropriate delays in repair. When 10 days was used as the cutoff for late repair, the complication rate was 14.4% (n = 88 of 610) for fractures repaired in less than 10 days vs 10.6% (n = 18 of 170) for the group treated in more than 10 days (P = .21). It is also apparent that the vast majority of fractures (n = 595 [76.3%]) were repaired between 3 and 10 days.
Delays in repair of maxillofacial injuries are often unavoidable. This study set out to identify the existing barriers to expeditious management of maxillofacial trauma in a tertiary care environment. Furthermore, we sought to evaluate the effect these delays had on complication rates. There is no motivation to suggest that medically unstable patients should undergo earlier fracture repair. Nor is there any thought that additional efforts or resources should be directed at optimizing patient compliance in the trauma population. Rather, in the age where an ever-increasing proportion of the care of these patients falls in the hands of the tertiary care facility, the concern, and thus our hypothesis, was that these demands for care may be overwhelming surgical resources.
The true effect of surgical delay in the management of maxillofacial fractures has been widely debated.2- 10,12,13 Although this study sought to examine the factors affecting the time to repair, the most appropriate measurement for optimal surgical outcomes with respect to time would appear to be rates of surgical complications. There are significant clinical concerns with respect to timing of repair that merit consideration. In situations where posttraumatic swelling may compromise the repair, an elective delay may be appropriate. Failure to allow posttraumatic edema to resolve may be a contributing factor to the statistically significant increase in complications performed prior to 2.5 days in our study, while later repair was equivocal. Later repairs of inherently contaminated wounds risk surgery in the setting of gross infection and its associated morbidity. In this study, surgical delays were quantifiable but generally not excessive. As such, this study may not be the best mechanism to measure the effect of true delay on morbidity. An observed complication rate of 13.6% is well within acceptable limits (Table 4).
In general, surgical delays can be described as the inability to achieve a confluence of surgeon, patient, and operating room. In the management of multisystem trauma, maxillofacial injuries generally receive a relatively low priority. In a previous study, Weider et al8 demonstrated that excessive soft tissue swelling, intracranial injuries, and unstable medical condition were the greatest causes for delay in their study group. In this study, variables associated with concomitant injures were most strongly associated with delayed repair. The GCS, a measure of neurologic status, is, in these circumstances, most likely indicative of associated neurologic injuries for which delay in repair is appropriate. Similarly, injury severity scores reflects more significant total body injury that may delay repair for similar reasons. Concomitant injuries may delay repair, however, for several reasons (Table 3). The most significant delay in repair is that the patient is too unstable as a result of these injuries to undergo maxillofacial surgery. However, it is not uncommon for a patient to require operations of variable urgency to address other sustained injuries that otherwise preclude the timely ability to address facial trauma. This is, to some degree, a logistical issue that is beyond the ability of this manuscript to address.
The significance of inpatient status at the time of surgery and its association with time to repair requires some mention (Table 3). At first glance, this would suggest that more severely injured patients, either from their facial injuries or their other injuries, remain admitted and may experience delayed repair for reasons noted herein. However, as many patients at our institution are many miles from home with limited resources for return visits, they may remain admitted until their surgery can be accomplished. One would expect, however, that given their inpatient status, their surgery should be accomplished in a timelier manner. The fact that this variable remains statistically significant in its association with delays in repair, especially when adjusting for variables associated with concomitant injuries, suggest an unmeasured logistical issue with “add-on” or “wait-list” surgery that is not detected by the limits of this study. We still conjecture that the logistical OR constraints remain undermeasured.
While issues related to patient scheduling and compliance were not statistically associated with observed delays in repair, it remains a nuisance in the management of the trauma patient. Patients presenting directly to our emergency department are the easiest to schedule and manage. The referral patterns at our institution, however, are such that many isolated maxillofacial injuries occur and are initially triaged at rural facilities many miles away. They are either accepted in transfer to our emergency department or referred to our clinic to be scheduled and managed as an outpatient. While the latter is cost-effective and reduces the burden on the emergency department, the former is unfortunately often more reliable. Despite nearly immediate clinic accessibility for these patients, surgeons must rely on the patient, their initiative to schedule the appointment, and their means to get there. It was surprising to observe this to be less of an issue in this cohort, as it frequently feels like a recurring struggle.
The inspiration for this article was the concern that potentially modifiable logistical factors were impeding the expeditious care of the maxillofacial trauma patient, but based on our data, this would not appear to be the case. Of the patients with a known delay, only 17 patients were identified with a delay directly attributable to the OR (Table 4), and these were cases where it was documented in the medical record that a patient either could not be scheduled for the OR or was scheduled and the case could not be accomplished. In an operating room that ordinarily runs at 85% capacity, it is not surprising that difficulties arise when scheduling these cases. As a level I trauma center, resources are available for the management of emergent trauma. As noted herein, maxillofacial injuries most frequently do not fit this description, as they are rarely emergent, and frequently given less consideration than one might give an exploratory laparotomy for penetrating trauma of the abdomen. Thus, maxillofacial injuries are worked in to the OR schedule almost electively, without the luxury of scheduling weeks in advance. The limitations of this study preclude the measurement of the following situations: (1) cases that were given a scheduled time much later than requested owing to no time available; (2) cases that were given a scheduled time after prior days on the waiting list; and (3) cases that were accomplished from the waiting list after several prior days on the waiting list. As a result, logistical issues resulting in treatment delay may remain underreported.
The retrospective nature of the study design requires reliance on the medical records for precluding the measurement of the 3 aforementioned situations and represents one of the most significant limitations of this study. While our retrospective approach allowed for completeness with regards to dates and times, as well as extraction of data from a meticulously kept hospital trauma database, the offline scheduling, rescheduling, moving, and delays of cases cannot be easily ascertained. Prospectively studying these scheduling delays deserves consideration; however, operating room scheduling and increased efficiency is a constant focus of quality improvement at our institution. Recent attempts to improve management of the patient population include specific multiservice block time for maxillofacial trauma during the week and heightened consideration for working in these cases, even as outpatients.
Multiple issues are associated with delays in repair of maxillofacial trauma. Most of these are associated with delays inherent to the multiple-trauma patient. Modifiable institutional logistical factors play a smaller role than anticipated but are likely underreported. There does not appear to be an association between delayed repair and the development of complications.
Corresponding Author: J. Paul Radabaugh, MD, Department of Otolaryngology, Head and Neck Surgery, University of Kentucky College of Medicine, 800 Rose St C236, Lexington, KY 40536 (email@example.com).
Published Online: January 14, 2016. doi:10.1001/jamafacial.2015.2101.
Author Contributions: Drs Radabaugh and Gal had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Radabaugh, Gal.
Acquisition, analysis, or interpretation of data: Radabaugh, Zhang, Wang, Lin, Shelton, Liau, Cunningham, Gal.
Drafting of the manuscript: Radabaugh, Lin, Shelton, Gal.
Critical revision of the manuscript for important intellectual content: Radabaugh, Zhang, Wang, Liau, Cunningham, Gal.
Statistical analysis: Lin, Gal.
Administrative, technical, or material support: Radabaugh, Zhang, Shelton, Cunningham, Gal.
Study supervision: Radabaugh, Liau, Gal.
Conflict of Interest Disclosures: None reported.
Additional Contributions: We thank David Akers, MS, and the Center for Clinical and Translational Science at the University of Kentucky for help with data analysis for this project. Mr Akers was not compensated for his contributions to this article.