Abstract
Importance: Time to surgical intervention is an oft-investigated potentially modifiable risk factor for complications after mandible fracture.
Objective: To identify novel risk factors for malunion/nonunion after mandible fracture and determine the impact of treatment delay on malunion and nonunion after open reduction of mandible fractures.
Design: Retrospective cohort.
Setting and Participants: Encounter billing records from the New York State Inpatient Databases, State Emergency Department Databases, and State Ambulatory Surgery Databases. Patients aged 18 years and older with isolated mandible fracture in the emergency department or inpatient setting from January 1, 2006 to September 30, 2015.
Main Outcomes and Measures: Mandibular Malunion/Nonunion.
Results: A total of 19,152 adults were diagnosed with isolated mandible fracture. After fracture, 247 patients (1.3%) developed mandibular malunion or nonunion. In multivariable analysis, patients with open fractures (odds ratio [OR] 1.93, confidence interval [95% CI] 1.40–2.65), body fractures (OR 2.00, 1.50–2.65), alcohol abuse (OR 1.61, 1.22–2.11), diabetes mellitus (OR 1.57, 1.02–2.42), and Medicaid insurance (OR 1.46, 1.03–2.07) had increased risk, whereas patients with subcondylar fractures had reduced risk (OR 0.45, 0.28–0.72) of mandibular malunion/nonunion. The risk of mandibular malunion/nonunion after open reduction increased with treatment delay until 6–7 days after presentation (OR 1.84, 1.11–3.06).
Conclusion and Relevance: Although treatment delay is often unavoidable, these findings suggest that physicians should consider early intervention in patients requiring open reduction of mandible fractures when able.
Key Points
Question: Does waiting to surgically repair a mandible fracture increase the risk of complications?
Findings: Patients who waited 6–7 days to have their mandible fracture surgically repaired had an increased risk of the bone not healing or healing poorly. Patients with diabetes, alcohol use disorder, Medicaid insurance, and fractures breaking the skin or in specific locations of the mandible also had a higher risk of bone-healing complications.
Meaning: When it is appropriate, physicians should consider surgery for mandible fractures sooner rather than later.
Introduction
Mandible fractures are the most common facial fractures requiring reduction in the United States with >10,000 inpatient cases per year.1 The majority of these reductions occur in young male patients who are uninsured or underinsured.2 Reductions are costly, averaging $26,035 and $49,892 for closed and open reduction, respectively,1 and serious complications of reduction, such as mandibular malunion and nonunion, often necessitate repeat operations.3,4 Many known risk factors for complications after fracture repair, such as mandibular body fracture,5 alcoholism,6,7 smoking,7–9 and patient comorbidities,9 are not modifiable.
Time to surgical repair after mandible fracture is an oft-investigated potentially modifiable risk factor for complications, but current literature on the topic is inconclusive. Although the majority of studies have not found any association between treatment delay and complication rates,5,6,8,10–13 some studies,14,15 including a recent study by Hsieh et al.,15 have found that treatment delays are associated with increased rates of complications. These studies are predominantly single-center retrospective studies of <300 patients with limited ability to conduct robust multivariable analyses and detect moderate effect sizes. The results of these studies are neither conducive to meta-analytic approaches, nor do they generalize well to the American adult population, as they are subject to substantial variability in patient demographics and physicians' practices.
Retrospective studies using administrative databases are an effective means of investigating risk factors for uncommon outcomes, due to their ability to obtain large sample sizes, include uncommon risk factors and complications, and capture a wide selection of patients from diverse demographic backgrounds. The goal of this study was to determine the impact of treatment delay on malunion and nonunion after open reduction of mandible fractures by utilizing the administrative databases provided by the Healthcare Cost and Utilization Project (HCUP). We also sought to identify novel risk factors for the development of malunion/nonunion after mandible fracture. We did not anticipate any meaningful association between the timing of open reduction and patients' risk of mandibular malunion/nonunion after treatment.
Methods
Study design
A retrospective cohort study was conducted utilizing the New York State Inpatient Databases (SID), State Emergency Department Databases (SEDD), and State Ambulatory Surgery Databases (SASD) from January 1, 2005 to December 31, 2016. This study was exempted from review by the Washington University School of Medicine institutional review board.
Data source
The SID,16 SEDD,17 and SASD18 are a family of databases developed by HCUP and maintained by the Agency for Healthcare Research and Quality. They contain state-specific files, including inpatient, emergency department visits not resulting in admission, and ambulatory surgery billing records. Their organizational structure and content are well suited for research that seeks to enumerate the number of discharges with specific diagnoses.
Study population
Patients aged 18 years and older with an ICD-9-CM diagnosis code for mandible fracture and no other ICD-9-CM diagnosis codes for fractures, except nasal fracture, in the SID or SEDD were included (Appendix Table A1). The data from SID and SASD for 2005 were used for identification of patient risk factors, and patients diagnosed with mandible fractures during this time were not included in the study population. Data from SEDD were only available from 2006 to 2016. Patients diagnosed with mandible fractures in the fourth quarter of 2015 and 2016 were not included in the study population to ensure a full year of follow-up data for all patients. The fourth quarter of 2015 was reserved for follow-up as it was the first quarter of ICD-10-CM codes. Patients with missing gender or personal identifiers were excluded, as were patients with out-of-state residence because of potential for loss to follow-up.
Patient demographics and comorbidities
Patient demographics, including gender, age, race, median household income of patient zip code (by quartile), and patient insurance, were evaluated. Patients were divided into age categories by quintile; elderly patients (≥65 years) were separated from the final quintile and evaluated separately. Owing to categories with small subsample size, race was collapsed into white, black, Hispanic, and other, and insurance was collapsed into Medicare, Medicaid, private, and other. Missing values for race and median household income were analyzed as separate categories as they are typically not missing at random. Patient comorbidities, or other medical ailments not related to mandible fracture, were defined by the Elixhauser system, a comprehensive set of comorbidity measures designed for used with large administrative data sets.19
Mandibular malunion/nonunion, mandible fracture treatment, and other fractures
The primary outcome of this study was malunion or nonunion after mandible fracture, defined as any ICD-9-CM or ICD-10-CM diagnosis codes for malunion or nonunion within 1 year of treatment (Appendix Table A2). If a patient had a procedure code for open reduction of the mandible within 21 days of diagnosis, the date of that procedure code was determined to be the date of definitive treatment (Appendix Table A3). If procedure codes existed on different dates, the earliest date was used. For patients who did not have open reduction, a procedure code for closed reduction within 21 days of diagnosis was considered the date of definitive treatment. If no procedures codes for treatment existed within the first 21 days, the patient was followed for 1 year after fracture diagnosis. Owing to categories with small subsample size, patients definitively treated on day 4 or 5 after diagnosis were grouped together; patients treated on day 6 or 7 were grouped together, and patients treated >1 week after diagnosis were grouped together.
Risk factors for infectious complications
The following risk factors for mandibular malunion/nonunion were evaluated: open fracture, fracture location, multiple fractures, edentulism, inflammatory jaw disease, jaw cysts, osteoporosis, firearm-related injury, and homelessness (Appendix Tables A1–A4).
Statistical analysis
Descriptive statistics were used to explore the distributions of patient demographics, comorbidities, fracture characteristics, and risk factors for mandibular malunion/nonunion. The association of patient characteristics and risk factors for mandibular malunion/nonunion was explored through generalized linear mixed models. Multivariable mixed models using variables significant in univariable analysis were used to identify independent risk factors for mandibular malunion/nonunion. The SAS PROC GENMOD procedure was used for all mixed models, accounting for nesting of patients within hospitals. The alpha level was set at 0.05 given the relatively small number of patients with certain risk factors. Odds ratios (OR) and 95% confidence intervals (95% CIs) were reported for all regression models. All statistical analysis was performed in SAS version 9.4 (Cary, NC). The relationship between age and mandibular malunion/nonunion was further explored through restricted cubic splines regression using Frank Harrell's SAS macro, PSPLINET.20 In cubic splines regression, “knots” are set at fixed points in the data (e.g., 5th, 35th, 65th, and 95th percentiles of age), and cubic splines (i.e., polynomials of degree 3) are used to model the risk of malunion/nonunion by age between knots. In restricted cubic splines regression, splines are constrained to be linear in the two tails.
Results
Patient demographics and comorbidities
Between January 1, 2006 and September 30, 2015, 19,152 adults were diagnosed with isolated mandible fracture(s) in the emergency department or inpatient setting in the state of New York (Table 1). Patients were predominantly male (N = 15,916, 83%) with median age at diagnosis of 28 (IQR, 22–42) years. Patients were typically white (N = 7239, 38%), from ZIP codes with median household incomes in the first quartile of the state (N = 7186, 38%), and insured by Medicaid (N = 6211, 32%). The prevalence of select comorbidities in our cohort is shown in Table 2. The most common comorbidities were alcohol abuse (N = 2993, 16%), drug abuse (N = 2410, 13%), chronic pulmonary disease (N = 1714, 9%), and psychoses (N = 1100, 6%).
Table 1.
Demographics of patients with isolated mandible fracture, NY, 2006–2015
| All patients (N = 19,152) |
Patients treated with open reduction (N = 7757) |
|||||
|---|---|---|---|---|---|---|
| Malunion/nonunion, N (%) | Total, N (%) | Malunion/nonunion, OR (95% CI) | Malunion/nonunion, N (%) | N (%) | Malunion/nonunion, OR (95% CI) | |
| Male | 206 (83) | 15,916 (83) | Ref. | 135 (84) | 6857 (88) | Ref. |
| Female | 41 (17) | 3236 (17) | 0.98 (0.68–1.41) | 26 (16) | 900 (12) | 1.48 (0.96–2.29) |
| Age | ||||||
| 18–20 | 17 (7) | 3346 (17) | Ref. | <11 | 1375 (18) | Ref. |
| 21–24 | 35 (14) | 3851 (20) | 1.79 (0.98–3.27) | 24 (15) | 1621 (21) | 2.91 (1.24–6.78) |
| 25–31 | 52 (21) | 4102 (21) | 2.49 (1.45–4.29) | 34 (21) | 1834 (23) | 3.64 (1.58–8.36) |
| 32–44 | 85 (34) | 3809 (20) | 4.37 (2.56–7.46) | 52 (32) | 1639 (21) | 6.23 (2.84–13.66) |
| 45–64 | ∼50 (21) | 2959 (15) | 3.38 (1.80–6.33) | 38 (24) | 1110 (14) | 6.73 (3.00–15.06) |
| 65+ | <11 | 1085 (6) | 1.26 (0.49–3.30) | <11 | 178 (2) | 6.62 (2.10–20.83) |
| Race | ||||||
| White | 96 (39) | 7239 (38) | Ref. | 60 (37) | 2208 (28) | Ref. |
| Black | 84 (34) | 6557 (34) | 0.96 (0.71–1.32) | 56 (35) | 3029 (39) | 0.68 (0.46–1.02) |
| Hispanic | 34 (14) | 2738 (14) | 0.94 (0.64–1.37) | 19 (12) | 1267 (16) | 0.55 (0.33–0.93) |
| Other | ∼30 (13) | 2470 (13) | 0.94 (0.67–1.32) | ∼30 (16) | 1194 (15) | 0.80 (0.55–1.17) |
| Missing | <11 | 148 (1) | 1.02 (0.24–4.34) | <11 | 59 (1) | * |
| Income | ||||||
| Fourth quartile | 32 (13) | 2700 (14) | Ref. | 24 (15) | 882 (11) | Ref. |
| Third quartile | 49 (20) | 3778 (20) | 1.09 (0.74–1.60) | 27 (17) | 1422 (18) | 0.69 (0.42–1.12) |
| Second quartile | 55 (22) | 4142 (22) | 1.12 (0.80–1.56) | 29 (18) | 1445 (19) | 0.72 (0.48–1.10) |
| First quartile | 83 (34) | 7186 (38) | 0.97 (0.68–1.38) | 58 (36) | 3117 (40) | 0.69 (0.45–1.04) |
| Missing | 28 (11) | 1346 (7) | 1.78 (1.04–3.06) | 23 (14) | 891 (11) | 0.98 (0.48–1.99) |
| Insurance | ||||||
| Medicare | 14 (6) | 1386 (7) | 1.12 (0.54–2.31) | <11 | 343 (4) | 1.83 (0.79–4.26) |
| Medicaid | 112 (45) | 6211 (32) | 2.00 (1.39–2.87) | 78 (48) | 3163 (40) | 1.57 (1.01–2.45) |
| Private | 49 (20) | 5371 (28) | Ref. | 30 (19) | 1835 (24) | Ref. |
| Self-pay | 57 (23) | 4785 (25) | 1.31 (0.85–2.03) | 32 (20) | 1687 (22) | 1.19 (0.65–2.17) |
| No charge | <11 | 135 (1) | * | <11 | 101 (1) | * |
| Other | 15 (6) | 1237 (6) | 1.27 (0.76–2.11) | 11 (7) | 1237 (8) | 0.98 (0.50–1.90) |
| Missing | <11 | 20 (0) | * | <11 | <11 (0) | * |
Bold values are statistically significant (p < 0.05).
Indicate sample size too small to perform analysis.
CI, confidence interval; OR, odds ratio.
Table 2.
Select Elixhauser comorbidities, characteristics of mandible fracture, and select risk factors in the population of patients with isolated mandible fracture, NY, 2006–2015
| All patients (N = 19,152) |
Open reduction (N = 7,757) |
|||||
|---|---|---|---|---|---|---|
| Malunion/nonunion, N (%) | Total, N (%) | Malunion/nonunion, OR (95% CI) | Malunion/nonunion, N (%) | Total, N (%) | Malunion/nonunion, OR (95% CI) | |
| Alcohol abuse | 72 (29) | 2993 (16) | 2.22 (1.69–2.92) | 51 (32) | 1368 (18) | 2.17 (1.60–2.95) |
| Deficiency anemias | 13 (5) | 521 (3) | 1.99 (1.17–3.37) | <11 | 179 (2) | 2.51 (1.42–4.43) |
| Chronic pulmonary disease | 32 (13) | 1714 (9) | 1.51 (1.03–2.23) | 24 (15) | 801 (10) | 1.53 (0.97–2.40) |
| Depression | 21 (9) | 1017 (5) | 1.66 (1.02–2.70) | 15 (9) | 407 (5) | 1.85 (1.09–3.14) |
| Diabetes mellitus | 18 (7) | 731 (4) | 1.99 (1.23–3.21) | 13 (8) | 282 (4) | 2.34 (1.46–3.76) |
| Drug abuse | 49 (20) | 2410 (13) | 1.72 (1.24–2.38) | 39 (24) | 1289 (17) | 1.62 (1.17–2.23) |
| Liver disease | <11 | 270 (1) | 2.03 (1.00–4.14) | <11 | 117 (2) | 1.65 (0.67–4.09) |
| Psychoses | 28 (11) | 1100 (6) | 2.10 (1.36–3.24) | 21 (13) | 480 (6) | 2.27 (1.31–3.96) |
| Renal failure | <11 | 180 (1) | 2.18 (0.96–4.96) | <11 | 51 (1) | 3.86 (1.49–9.97) |
| Fracture | ||||||
| Closed | 196 (79) | 17,203 (90) | Ref. | 120 (75) | 6454 (83) | Ref |
| Open | 51 (21) | 1949 (10) | 2.30 (1.66–3.18) | 41 (25) | 1303 (17) | 1.71 (1.17–2.48) |
| Location | ||||||
| Alveolar border | <11 | 361 (2) | 0.42 (0.10–1.78) | <11 | 85 (1) | 0.57 (0.08–4.18) |
| Symphyseal | 38 (15) | 2706 (14) | 1.10 (0.77–1.56) | 30 (19) | 2186 (28) | 0.58 (0.39–0.87) |
| Body | 64 (26) | 2349 (12) | 2.50 (1.92–3.26) | 51 (32) | 1499 (19) | 1.94 (1.39–2.70) |
| Angle | 72 (29) | 4472 (23) | 1.35 (1.01–1.79) | 56 (35) | 2967 (38) | 0.87 (0.64–1.17) |
| Ramus | 27 (11) | 1758 (9) | 1.21 (0.77–1.91) | 21 (13) | 638 (8) | 1.67 (1.00–2.78) |
| Subcondylar | 15 (6) | 1877 (10) | 0.59 (0.35–0.97) | 14 (9) | 1012 (13) | 0.64 (0.38–1.06) |
| Condylar | 11 (4) | 1345 (7) | 0.61 (0.36–1.05) | <11 | 305 (4) | 0.94 (0.47–1.86) |
| Coronoid | <11 | 117 (1) | 0.66 (0.10–4.27) | <11 | 36 (1) | 1.34 (0.20–9.17) |
| Multiple | 89 (36) | 4850 (25) | 1.66 (1.29–2.14) | 70 (43) | 3460 (45) | 0.95 (0.70–1.31) |
| Unspecified | 64 (26) | 6733 (35) | 0.65 (0.50–0.84) | 28 (17) | 1385 (18) | 0.96 (0.67–1.39) |
| Edentulous | <11 | 83 (1) | 0.93 (0.13–6.59) | <11 | 40 (1) | 1.27 (0.16–9.03) |
| Inflammatory jaw disease | <11 | 55 (0) | 4.28 (1.42–12.93) | <11 | 20 (0) | 4.90 (1.26–18.97) |
| Jaw cysts | <11 | 42 (0) | 1.84 (0.33–10.36) | <11 | 33 (0) | 1.45 (0.28–7.49) |
| Osteoporosis | <11 | 113 (1) | 1.38 (0.37–5.20) | <11 | 28 (0) | 3.49 (1.11–10.96) |
| Firearm-related injury | <11 | 151 (1) | 2.60 (1.18–5.73) | <11 | 64 (1) | 1.52 (0.44–5.20) |
| Homeless | 31 (13) | 1334 (7) | 2.00 (1.39–2.86) | 26 (16) | 859 (11) | 1.64 (1.00–2.67) |
Bold values are statistically significant (p < 0.05).
Fracture characteristics and treatment
Most patients with isolated mandible fractures had closed fractures (N = 17,203, 90%). Mandible fractures commonly occurred at the angle (N = 4472, 23%), symphysis (N = 2706, 14%), and body (N = 2349, 12%) of the mandible. Fracture location was not specified in 35% of patients (N = 6733), and 25% of patients (N = 4850) had multiple mandibular fractures (Table 2).
Of the 19,152 patients with mandible fractures, 7796 patients (41%) were treated with open reduction, 2430 patients (13%) were treated with closed reduction, and 69 patients (0.4%) were treated with other open procedures. The remaining 8875 patients (46%) did not have a procedure code for treatment recorded within the first 3 weeks of fracture diagnosis. Among patients who underwent open reduction, median time to treatment was 2 (IQR, 1–4) days. Most patients with mandible fractures treated with open reduction had surgery within 7 days of diagnosis (N = 7264, 94%).
Incidence and timing of mandibular malunion/nonunion
In total, 247 patients (1.3%) developed mandibular malunion (N = 79, 0.4%) and/or nonunion (N = 181, 1.0%) after mandible fracture. The median time to diagnosis of mandibular malunion/nonunion after treatment was 60 (IQR, 30–120) days and 68 (IQR, 43–104) days, respectively. One-year rates of mandibular malunion/nonunion were higher among patients treated with open reduction (N = 161, 2.1%) than patients treated with closed reduction (N = 29, 1.2%) or patients with no record of treatment (N = 56, 0.6%).
Risk factors associated with mandibular malunion/nonunion in all patients
The associations of patient demographics, comorbidities, and fracture characteristics with mandibular malunion/nonunion when controlling for other risk factors in multivariable analysis are shown in Table 3.
Table 3.
Multivariable analysis, association of risk factors with mandibular malunion and nonunion in patients with isolated mandible fracture, NY, 2006–2015
| Malunion/nonunion, N (%) | Total, N (%) | Malunion/nonunion, OR (95% CI) | |
|---|---|---|---|
| Age | |||
| 18–20 | 17 (7) | 3346 (17) | Ref. |
| 21–24 | 35 (14) | 3851 (20) | 1.72 (0.94–3.14) |
| 25–31 | 52 (21) | 4102 (21) | 2.26 (1.32–3.89) |
| 32–44 | 85 (34) | 3809 (20) | 3.94 (2.31–6.70) |
| 45–64 | ∼50 (21) | 2959 (15) | 2.84 (1.50–5.40) |
| 65+ | <11 | 1085 (6) | 1.51 (0.54–4.27) |
| Fracture | |||
| Closed | 196 (79) | 17,203 (90) | Ref. |
| Open | 51 (21) | 1949 (10) | 1.93 (1.40–2.65) |
| Location | |||
| Body | 64 (26) | 2349 (12) | 2.00 (1.50–2.65) |
| Subcondylar | 15 (6) | 1877 (10) | 0.45 (0.28–0.72) |
| Alcohol abuse | 72 (29) | 2993 (16) | 1.61 (1.22–2.11) |
| Diabetes mellitus | 18 (7) | 731 (4) | 1.57 (1.02–2.42) |
| Insurance | |||
| Medicare | 14 (6) | 1386 (7) | 1.02 (0.48–2.16) |
| Medicaid | 112 (45) | 6211 (32) | 1.46 (1.03–2.07) |
| Private | 49 (20) | 5371 (28) | Ref. |
| Other | 72 (29) | 72 (32) | 1.13 (0.77–1.66) |
Bold values are statistically significant (p < 0.05).
Patients aged 25–64 years were more than twice as likely to develop mandibular malunion/nonunion when compared with young adults aged 18–20 years. Similarly, patients with open fractures (OR 1.93, 95% CI 1.40–2.65) and body fractures (OR 2.00, 95% CI 1.50–2.65) had increased risk of mandibular malunion/nonunion, whereas patients with subcondylar fractures had reduced risk of malunion/nonunion (OR 0.45, 0.28–0.72). Patients with a history of alcohol abuse (OR 1.61, 1.22–2.11) or diabetes mellitus (OR 1.57, 1.02–2.42) had increased likelihood of mandibular malunion/nonunion. Compared with patients with private insurance, patients insured by Medicaid were at increased risk of mandibular malunion/nonunion (OR 1.46, 1.03–2.07).
Risk factors associated with mandibular malunion/nonunion after open reduction
The associations of patient demographics, comorbidities, and fracture characteristics with mandibular malunion/union in multivariable analysis for patients treated with open reduction are shown in Table 4.
Table 4.
Multivariable analysis, association of risk factors with mandibular malunion and nonunion in patients with isolated mandible fracture treated with open reduction, NY, 2006–2015
| Malunion/nonunion, N (%) | Total, N (%) | Malunion/nonunion, OR (95% CI) | |
|---|---|---|---|
| Age | |||
| 18–20 | <11 | 1375 (18) | Ref. |
| 21–24 | 24 (15) | 1621 (21) | 2.76 (1.19–6.42) |
| 25–31 | 34 (21) | 1834 (23) | 3.27 (1.43–7.48) |
| 32–44 | 52 (32) | 1639 (21) | 5.44 (2.47–11.98) |
| 45–64 | 38 (24) | 1110 (14) | 5.20 (2.29–11.78) |
| 65+ | <11 | 178 (2) | 5.47 (1.78–16.83) |
| Open reduction timing (days after presentation) | |||
| 0 | 23 (14) | 1487 (19) | Ref. |
| 1 | ∼40 (24) | 1937 (25) | 1.35 (0.83–2.18) |
| 2 | 29 (18) | 1413 (18) | 1.26 (0.78–2.04) |
| 3 | 22 (14) | 945 (12) | 1.44 (0.78–2.65) |
| 4–5 | 25 (16) | 1028 (13) | 1.50 (0.90–2.49) |
| 6–7 | 14 (9) | 454 (6) | 1.84 (1.11–3.06) |
| 7+ | <11 | 493 (6) | 1.13 (0.57–2.24) |
| Fracture | |||
| Closed | 120 (75) | 6454 (83) | Ref. |
| Open | 41 (25) | 1303 (17) | 1.72 (1.21–2.44) |
| Location | |||
| Body | 51 (32) | 1499 (19) | 1.73 (1.25–2.39) |
| Subcondylar | 14 (9) | 1012 (13) | 0.57 (0.35–0.93) |
| Alcohol abuse | 51 (32) | 1368 (18) | 1.70 (1.24–2.33) |
| Diabetes mellitus | 13 (8) | 282 (4) | 1.52 (0.95–2.42) |
Bold values are statistically significant (p < 0.05).
Among patients treated with open reduction, adults aged 21 years and older had increased risk of mandibular malunion/nonunion when compared with young adults aged 18–20 years; the effect was most pronounced in patients aged 32 years and older, who experienced a fivefold increase in risk. The relationship between patient age and risk of malunion/nonunion, in all patients and specifically in patients treated with open reduction, is further explored with restricted cubic splines in Figure 1. Among patients receiving all forms of treatment, risk of malunion/nonunion decreased after the age of 40 years; however, among patient treated with open reduction, they remained elevated in patients >40 years.
Fig. 1.
The risk of malunion or nonunion by age in patients with isolated mandible fractures and in patients with isolated mandible fractures treated with open reduction are shown in (A, B), respectively.
The odds of mandibular malunion/nonunion after open reduction gradually increased with treatment delay up until the sixth and seventh day after treatment (OR 1.84, 1.11–3.06). Patients treated with open reduction greater than a week after presentation did not have significantly elevated rates of malunion/nonunion (OR 1.13, 0.57–2.24) when compared with those who had surgery on the day of presentation.
Discussion
In this study, we observed a dose–response trend with increasing risk of mandibular malunion or nonunion for each day that open reduction is delayed for up to 1 week after diagnosis. These findings are consistent with a recent study by Hsieh et al.,15 in which patients who developed complications after mandible fracture had a mean time to surgical treatment of 5.9 days. As mandibular malunion/nonunion is an uncommon outcome, previous studies5,6,8,10–13 that did not show any association between treatment timing and complication rates, were likely limited by small sample sizes and underpowered, at least with regard to detecting differences in rates of malunion/nonunion. Although the precise timing for ideal treatment is not known, and many factors may preclude early surgical intervention, including comorbid patient conditions or marked facial swelling, physicians should consider early intervention for patients with mandible fractures requiring open reduction when able.
Patients who waited at least 1 week for open reduction did not have increased risk of malunion/nonunion. There are many possible explanations for this observation. First, only 6% of open reductions in this study occurred after 1 week, so the estimates of risk associated during this timeframe are imprecise. Second, other complications, such as infection, which are not accounted for in this study may predominate in patients treated during this timeframe. Finally, delayed treatment greater than a week after presentation may be reflective of patients with poor access to care; patients with poor access to care are inherently less likely to present for follow-up visits and seek treatment for complications such as nonunion/malunion after fracture repair.
This study highlights age as important risk factor for malunion/nonunion after mandible fracture. For patients receiving all types of treatment, age 25–64 years was associated with increased risk of malunion/nonunion after mandible fracture. In patients who underwent open reduction, elderly patients (65+) were also at increased risk for malunion/nonunion. These findings likely reflect the quality of bone and mechanism of injury common to these demographics. Elderly patients are more likely to have fragility fractures, or mandible fracture secondary to falls in the setting of atrophic mandibles related to age, poor dentition, or edentulous mandibles. Younger adults are more likely to have more severely comminuted or displaced mandible fractures from high-impact mechanisms such as motor vehicle collisions, gunshot wounds, or assault.21 However, elderly patients with mandible fractures severe enough to warrant surgery had approximately the same risk of malunion/nonunion as middle-aged adults.
To our knowledge, this is the first study to use the SID, SASD, and SEDD to examine the impact of treatment delay on mandible fracture and identify risk factors for malunion/nonunion after mandible fracture. In contrast to previous studies, it benefits from a large sample size conducive to multivariable analysis allowing comparison of the relative contributions of various risk factors with rates of mandibular malunion/nonunion.
The limitations of this study are similar to those of any administrative database study, which rely upon medical coders to accurately capture risk factors and medical complications. Minor diagnoses that do not generate revenue are often not consistently coded. This limitation could explain the low number of specific risk factors (e.g., osteoporosis, edentulism, and jaw cysts). This study was also only able to study one type of complication after mandible fracture, nonunion/malunion, and would have benefitted from more explicit ICD-9/10-CM coding for infectious complications; for example, osteomyelitis is coded under inflammatory conditions of the jaws, which is a broad code including chronic and unrelated conditions such as hard palate abscesses and giant cell hyaline angiopathy. Other limitations of this study include inability to classify fractures as favorable or unfavorable, to delineate which specific reduction techniques were used, or to control for repeat trauma to the mandible after open reduction.
Conclusion
This study finds an increasing risk of mandibular malunion/nonunion for patients treated with open reduction as treatment is delayed. Although treatment delay is often unavoidable, these findings suggest that physicians should consider early intervention in patients requiring open reduction of mandible fractures when able. Future studies to investigate optimal timing of open reduction and evaluate the impact of treatment delay on other complications of mandible fracture are warranted.
Appendix
Appendix Table A1. ICD-9 codes for defining the index cohort
| ICD-9 codes | Description |
|---|---|
| 80220, 80221, 80222, 80223, 80224, 80225, 80226, 80227, 80228, 80229 80230, 80231, 80232, 80233, 80234, 80235, 80236, 80237, 80238, 80239 |
Mandible fracture |
Appendix Table A2. ICD-9/10 codes for defining primary outcome
| ICD-9/10 codes | Description |
|---|---|
| 73381 S026**G |
Malunion |
| 73382 S026**K |
Nonunion |
Appendix Table A3. ICD-9, CPT-4 codes for open reduction
| ICD-9, CPT-4 codes | Description |
|---|---|
| 7676 21445, 21461, 21462, 21465, 21470 |
Open reduction |
Appendix Table A4. ICD-9 codes for defining risk factors
| ICD-9 | Description |
|---|---|
| 5254, 5255 | Edentulous |
| 5264 | Inflammatory jaw disease |
| 5260, 5261, 5262 | Jaw cysts |
| 7330 | Osteoporosis |
| E922, E955, E965, E970, E985, E991 | Firearm-related injury |
Authors' Contributions
Design, conduct, analysis, and presentation of the research by D.P.L. and M.A.O. Design, analysis, and presentation of the research by J.J.L., D.K., J.F.P., and E.A.S. Conduct and analysis by D.S. All coauthors have reviewed and approved of the article before submission.
Author Disclosure Statement
No competing financial interests exist.
Funding Information
This study was supported by the T32 DC00022 from the National Institutes of Deafness and Other Communication Disorders (NIDCD); the Center for Administrative Data Research (CADR), Washington University Institute of Clinical and Translational Sciences (ICTS) supported by UL1TR002345 from the NIH/National Center for Advancing Translational Sciences (NCATS), and R24HS19455 from the Agency for Healthcare Research and Quality (AHRQ). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
References
- 1. Nalliah RP, Allareddy V, Kim MK, et al. Economics of facial fracture reductions in the United States over 12 months. Dent Traumatol. 2013;29(2):115–120. [DOI] [PubMed] [Google Scholar]
- 2. Pena I Jr., Roberts LE, Guy WM, et al. The cost and inpatient burden of treating mandible fractures: a nationwide inpatient sample database analysis. Otolaryngol Head Neck Surg. 2014;151(4):591–598. [DOI] [PubMed] [Google Scholar]
- 3. Ostrander BT, Wang HD, Cusano A, et al. Contemporary management of mandibular fracture nonunion—a retrospective review and treatment algorithm. J Oral Maxillofac Surg. 2018;76(7):1479–1493. [DOI] [PubMed] [Google Scholar]
- 4. Vega LG. Reoperative mandibular trauma: management of posttraumatic mandibular deformities. Oral Maxillofac Surg Clin North Am. 2011;23(1):47–61, v–vi. [DOI] [PubMed] [Google Scholar]
- 5. Lee UK, Rojhani A, Herford AS, et al. Immediate versus delayed treatment of mandibular fractures: a stratified analysis of complications. J Oral Maxillofac Surg. 2016;74(6):1186–1196. [DOI] [PubMed] [Google Scholar]
- 6. Furr AM, Schweinfurth JM, May WL. Factors associated with long-term complications after repair of mandibular fractures. Laryngoscope. 2006;116(3):427–430. [DOI] [PubMed] [Google Scholar]
- 7. Serena-Gomez E, Passeri LA. Complications of mandible fractures related to substance abuse. J Oral Maxillofac Surg. 2008;66(10):2028–2034. [DOI] [PubMed] [Google Scholar]
- 8. Chen CL, Zenga J, Patel R, et al. Complications and reoperations in mandibular angle fractures. JAMA Facial Plast Surg. 2018;20(3):238–243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Gutta R, Tracy K, Johnson C, et al. Outcomes of mandible fracture treatment at an academic tertiary hospital: a 5-year analysis. J Oral Maxillofac Surg. 2014;72(3):550–558. [DOI] [PubMed] [Google Scholar]
- 10. Lucca M, Shastri K, McKenzie W, et al. Comparison of treatment outcomes associated with early versus late treatment of mandible fractures: a retrospective chart review and analysis. J Oral Maxillofac Surg. 2010;68(10):2484–2488. [DOI] [PubMed] [Google Scholar]
- 11. Barker DA, Oo KK, Allak A, et al. Timing for repair of mandible fractures. Laryngoscope. 2011;121(6):1160–1163. [DOI] [PubMed] [Google Scholar]
- 12. Odom EB, Snyder-Warwick AK. Mandible fracture complications and infection: the influence of demographics and modifiable factors. Plast Reconstr Surg. 2016;138(2):282e–289e. [DOI] [PubMed] [Google Scholar]
- 13. Stone N, Corneman A, Sandre AR, et al. Treatment delay impact on open reduction internal fixation of mandibular fractures: a systematic review. Plast Reconstr Surg Glob Open. 2018;6(6):e1829. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Biller JA, Pletcher SD, Goldberg AN, et al. Complications and the time to repair of mandible fractures. Laryngoscope. 2005;115(5):769–772. [DOI] [PubMed] [Google Scholar]
- 15. Hsieh TY, Funamura JL, Dedhia R, et al. Risk factors associated with complications after treatment of mandible fractures. JAMA Facial Plast Surg. 2019;21(3):213–220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. HCUP State Inpatient Databases (SID). Healthcare Cost and Utilization Project (HCUP). 2005–2016. [Google Scholar]
- 17. HCUP State Emergency Department Databases (SEDD). Healthcare Cost and Utilization Project (HCUP). 2005–2016. [Google Scholar]
- 18. HCUP State Ambulatory Surgery and Services Databases (SASD). Healthcare Cost and Utilization Project (HCUP). 2005–2016. [Google Scholar]
- 19. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27. [DOI] [PubMed] [Google Scholar]
- 20. Harrell F. SAS Macros for Assisting with Survival and Risk Analysis, and Some SAS Procedures Useful for Multivariable Modeling. 1999. [Google Scholar]
- 21. King RE, Scianna JM, Petruzzelli GJ. Mandible fracture patterns: a suburban trauma center experience. Am J Otolaryngol. 2004;25(5):301–307. [DOI] [PubMed] [Google Scholar]