Abstract
Objectives
Osteomyelitis of the long bones can result from hematogenous spread, direct inoculation or from a contiguous focus of infection. The association of osteomyelitis after long bone fractures has widely been believed to be true by practicing surgeons since the 1950s, even though the evidence has been poor. We hypothesized that long bone shaft fracture and major bone surgery are independent risk factors for osteomyelitis in adult trauma patients.
Methods
The National Trauma Data Bank (NTDB) was queried between 2007 and 2015 for patients ≥18 years of age presenting after trauma. Patients with long bone shaft fractures (femur, tibia/fibula, humerus) were identified and rate of acute osteomyelitis was calculated. Univariable logistic regression was performed. A multivariable logistic regression was performed to identify risk factors for development of acute osteomyelitis.
Results
From 5,494,609 patients, 358,406 were identified to have long bone shaft fractures (6.5%) with the majority being tibia/fibula (44.3%). The osteomyelitis rate in long bone shaft fractures was 0.05%. Independent risk factors for osteomyelitis included major humerus surgery and major tibia/fibula surgery. The strongest risk factor was non-pseudomonas bacteremia. Long bone shaft fractures were not found to be an independent risk factor for osteomyelitis (p > 0.05).
Conclusions
Long bone shaft fractures are not independently associated with increased risk for osteomyelitis. Major extremity surgery on the humerus and tibia/fibula, but not femur, are independent risk factors for osteomyelitis. However, the strongest risk factor is non-pseudomonas bacteremia.
Keywords: Osteomyelitis, Trauma, Long bone fracture, Bacteremia
1. Introduction
Dating back to Hippocrates, infection after bony fracture is one of the oldest diseases known to man.1 However, acute osteomyelitis was not described clinically until 1773.1 In traumatic cases, acute osteomyelitis has been reported to occur most commonly following open long bone fractures with an incidence of 4–63%.2 Acute osteomyelitis of the long bones can result from hematogenous spread, direct inoculation or from a contiguous focus of infection, with direct inoculation thought to be the primary cause in trauma patients.3,4
The rate of open long bone fractures has been reported at 11.5 per 100,000 persons per year, with osteomyelitis complicating up to 30% of cases.5,6 For this reason, surgical debridement and irrigation should occur as soon as possible. Within six hours of the trauma, the threshold density for open fractures, ≥ 105 organisms per gram of tissue, is reached.7 However, some studies have suggested that surgical intervention itself may be an independent risk factor for the development of osteomyelitis after open long bone fractures, while others concluded surgical intervention to be protective.8, 9, 10
Although the relationship between acute osteomyelitis and long bone fracture remains poorly understood, the risk of acute osteomyelitis after long bone fractures has not been studied using a large national trauma database. The purpose of this study was to determine risk factors for development of acute osteomyelitis in adult trauma patients presenting with long bone fracture. We hypothesized that long bone shaft fracture and major bone surgery are risk factors for acute osteomyelitis.
2. Methods
This was a retrospective analysis using the National Trauma Data Bank (NTDB) which consists of a multicenter registry of trauma centers in North America maintained by the American College of Surgeons Committee on Trauma.11 This study was deemed exempt by the institutional review board of the University of California, Irvine, as the NTDB is a national de-identified database. We queried the NTDB from January 2007 to December 2015 to identify all patients admitted with open or closed long bone (femur, tibia/fibula, humerus) shaft fractures using the International Classification of Diseases (ICD) version-9 diagnosis codes listed in Appendix A. Patients <18 years of age were excluded.
Our primary end-point of interest was the development of acute osteomyelitis during the index hospitalization. To evaluate which patients with long bone shaft fractures are associated with osteomyelitis, we compared patients with femur, humerus and tibia/fibula shaft fractures. In order to identify risk factors for developing osteomyelitis, we additionally compared patients with and without osteomyelitis.
Secondary outcomes included total hospital length of stay (LOS), intensive care unit (ICU) LOS, ventilator days, acute kidney injury (AKI), acute respiratory distress syndrome (ARDS), myocardial infarction (MI), pulmonary embolism (PE), pneumonia, unplanned intubation, unplanned ICU admission, pseudomonas and non/pseudomonas bacteremia, blood transfusion requirement and mortality. We also identified patients that underwent major bone surgery. For the purposes of this study, major bone surgery was defined as any surgery involving application of external fixation device, reconstructive operation, internal fixation without reduction, closed reduction with and without internal fixation, open reduction with and without internal fixation, debridement and unspecified bone surgery. The relationship between acute osteomyelitis and baseline patient demographics, comorbidities, injury profile, interventions and hospital outcomes were analyzed.
Patient demographic information including age, gender and pre-hospital comorbidities including diabetes, end stage renal disease (ESRD), steroid use, smoking and peripheral arterial disease (PAD) were collected. The injury profile included the injury severity score (ISS), abbreviated injury scale (AIS) for body region, blunt mechanism of injury, hypotension on arrival (systolic blood pressure ≤ 90 mmHg) and presence of positive blood alcohol concentration (BAC) or illegal drug screen on admission. Associated injuries were identified by the appropriate ICD-9 diagnosis codes and included spine injury, traumatic brain injury (TBI) and extremity fracture. All variables were coded as present or absent.
Descriptive statistics were performed for all variables. A Student's t-test was used to compare continuous variables and chi-square was used to compare categorical variables for bivariate analysis. Categorical data were reported as percentages, and continuous data were reported as medians with interquartile range or means with standard deviation.
The magnitude of the association between predictor variables and acute osteomyelitis was first measured using a univariable logistic regression model. We included previously reported risk factors for osteomyelitis into a univariable linear regression model.12, 13, 14, 15, 16, 17 Covariates with statistical significance (p ≤ 0.20) were included in a hierarchical multivariable logistic regression model and the adjusted risk for acute osteomyelitis was reported with an odds ratio (OR) and 95% confidence intervals (CI). The reference group used in our analysis comparing long bone shaft fractures included all adult patients with documented long bone shaft fractures. The reference group used in our analysis to determine risk for acute osteomyelitis included all adult patients in the dataset. All p-values were two-sided, with a statistical significance level of <0.05. All missing data points were not imputed but treated as missing data. All analyses were performed with IBM SPSS Statistics for Windows (Version 24, IBM Corp., Armonk, NY).
3. Results
3.1. Demographics of patients with long bone shaft fractures and primary outcome
From 5,494,609 adult trauma admissions between 2007 and 2015, 358,406 patients were identified to have long bone shaft fractures (6.5%), with the largest percentage being tibia/fibula (44.3%). The majority of long bone shaft fractures were closed (femur, 82.2%; tibia/fibula 69.4%; humerus, 82.3%, p < 0.001). Compared to all patients with long bone shaft fractures, those with femur shaft fractures were younger (mean age, 44.6 vs. 45.3, p < 0.001) with a higher median ISS (10.0 vs. 9.5, p < 0.001) and higher rate of TBI (20.5% vs. 19.0%, p < 0.001). Patients with tibia/fibula shaft fractures were younger (mean age, 43.7 vs. 45.3, p < 0.001) with a lower median ISS (9.0 vs. 9.5, p < 0.001) while those with humerus shaft fractures were older (mean age, 49.2 vs. 45.3, p < 0.001) with a lower median ISS (9.0 vs. 9.5, p < 0.001) and higher rate of TBI (24.8% vs. 19.0%, p < 0.001) (Table 1). Compared to patients with closed long bone shaft fractures, those with open fractures were younger (mean age, 39.0 vs. 47.1, p < 0.001), more likely to be male (76.5% vs. 59.7%) and less likely to have a blunt mechanism of injury (77.7% vs. 98.8%) (Table 2). The overall rate of acute osteomyelitis was 0.05%. Compared to all patients with long bone shaft fractures, those with femur (0.01% vs. 0.05%, p < 0.05) and humerus (0.03% vs. 0.05%, p < 0.001) shaft fractures had a lower rate of acute osteomyelitis while those with tibia/fibula shaft fractures had a higher rate of acute osteomyelitis (0.1% vs. 0.05%, p < 0.001) (Table 3). ISS and age were missing in 4.8% and 6.7% of patients; these patients were excluded from regression analysis.
Table 1.
Patient characteristics of all adult trauma patients with long bone shaft fractures.
| Characteristic | Long bone shaft fracture |
Femur |
Tibia/fibula |
Humerus |
|||
|---|---|---|---|---|---|---|---|
| (n = 358406) | (n = 146685) | p-value | (n = 158749) | p-value | (n = 69856) | p-value | |
| Age, year, mean (SD) | 45.3 (18) | 44.6 (22) | <0.001 | 43.7 (27) | <0.001 | 49.2 (21) | <0.001 |
| Sex, male, n (%) | 226305 (63.1) | 92313 (63.2) | <0.05 | 107996 (68.3) | <0.001 | 37588 (54.0) | <0.001 |
| Blunt mechanism, n (%) | 328507 (91.7) | 134550 (93.7) | <0.001 | 147105 (95.7) | <0.001 | 63127 (92.6) | <0.001 |
| ISS, median (IQR) | 9.5 (6) | 10.0 (10) | <0.001 | 9.0 (9) | <0.001 | 9.0 (9) | <0.001 |
| Shaft fracture, n (%) | |||||||
| Open | 82791 (23.1) | 26109 (17.8) | <0.001 | 48577 (30.6) | <0.001 | 12364 (17.7) | <0.001 |
| Closed | 287083 (80.1) | 120629 (82.2) | <0.001 | 110175 (69.4) | <0.001 | 57473 (82.3) | <0.001 |
| Injuries, n (%) | |||||||
| Traumatic brain injury | 68233 (19.0) | 30140 (20.5) | <0.001 | 27218 (17.1) | <0.001 | 17309 (24.8) | <0.001 |
| Spine | 56561 (15.8) | 24653 (16.8) | <0.001 | 23339 (14.7) | <0.001 | 14300 (20.5) | <0.001 |
| Upper extremity fracture | 121430 (33.9) | 34162 (23.3) | <0.001 | 28322 (17.8) | <0.001 | 69856 (100) | – |
| Lower extremity fracture | 304977 (85.1) | 146685 (100) | – | 158749 (100) | – | 16427 (23.5) | <0.001 |
| AIS (severe)a, n (%) | |||||||
| Head | 23058 (6.4) | 10394 (7.1) | <0.001 | 8876 (5.6) | <0.001 | 6300 (9.0) | <0.001 |
| Spine | 1586 (0.4) | 670 (0.5) | <0.001 | 596 (0.4) | <0.001 | 462 (0.7) | <0.001 |
| Thorax | 16537 (4.6) | 7798 (5.3) | <0.001 | 5890 (3.7) | <0.001 | 4861 (7.0) | <0.001 |
| Abdomen | 7365 (2.1) | 3973 (2.7) | <0.001 | 2336 (1.5) | <0.001 | 2094 (3.0) | <0.001 |
SD = standard deviation, IQR = interquartile range, AIS = abbreviated injury scale.
= severe (grade>3).
Table 2.
Patient characteristics of all adult trauma patients with open and closed long bone fractures.
| Characteristic | Closed |
Open |
p-value |
|---|---|---|---|
| (n = 279375) | (n = 79031) | ||
| Age, years, mean (SD) | 47.1 (20.6) | 39.0 (16.7) | <0.001 |
| Sex, male, n (%) | 166065 (59.7) | 60240 (76.5) | <0.001 |
| Blunt mechanism, n (%) | 268618 (98.8) | 59889 (77.7) | <0.001 |
| Injuries, n (%) | |||
| Traumatic brain injury | 52942 (19.0) | 15291 (19.3) | <0.05 |
| Spine | 43998 (15.7) | 12563 (15.9) | 0.32 |
| Isolated upper extremity fracture | 96161 (34.4) | 25269 (32.0) | <0.001 |
| Isolated lower extremity fracture | 235399 (84.3) | 69578 (88.0) | <0.001 |
| AIS (severe)a, n (%) | |||
| Head | 17756 (6.4) | 5302 (6.7) | <0.001 |
| Spine | 1133 (0.4) | 453 (0.6) | <0.001 |
| Abdomen | 5257 (1.9) | 2108 (2.7) | <0.001 |
| Thorax | 12406 (4.4) | 4131 (5.2) | <0.001 |
| LOS, days, mean (SD) | 8.1 (10.5) | 10.1 (12.7) | <0.001 |
| ICU, days, mean (SD) | 7.7 (9.3) | 7.7 (9.6) | <0.05 |
| Ventilator, days, mean (SD) | 7.82 (10.2) | 6.9 (9.2) | <0.001 |
SD = standard deviation, AIS = abbreviated injury scale, LOS = length of stay, ICU = intensive care unit.
= severe (grade>3).
Table 3.
Outcomes of all adult trauma patients with long bone shaft fractures.
| Characteristic | All |
Femur |
Tibia/fibula |
Humerus |
|||
|---|---|---|---|---|---|---|---|
| (n = 358406) | (n = 146685) | p-value | (n = 158749) | p-value | (n = 69856) | p-value | |
| Outcomes | |||||||
| LOS, days, mean (SD) | 8.5 (11) | 9.3 (11) | <0.001 | 8.6 (11) | <0.001 | 8.5 (12) | <0.001 |
| ICU, days, mean (SD) | 7.7 (11) | 7.7 (9) | 0.23 | 7.8 (10) | <0.001 | 7.5 (9) | <0.001 |
| Ventilator, days, mean (SD) | 7.6 (6) | 4.0 (8) | <0.05 | 7.7 (10) | <0.001 | 8.1 (10) | <0.001 |
| Major surgery, n (%) | |||||||
| All | 178486 (49.8) | 116047 (79.1) | <0.001 | 117550 (74.0) | <0.001 | 36948 (52.9) | <0.001 |
| Application of external fixation device | 35123 (9.8) | 9109 (6.2) | <0.001 | 23729 (14.9) | <0.001 | 1173 (1.7) | <0.001 |
| Reconstructive operation | 1433 (0.4) | 251 (0.2) | <0.001 | 962 (0.6) | <0.001 | 65 (0.1) | <0.001 |
| Internal fixation without reduction | 15053 (4.2) | 11599 (3.2) | <0.001 | 8960 (5.6) | <0.001 | 924 (1.3) | <0.001 |
| Closed reduction without internal fixation | 22937 (6.4) | 8700 (5.9) | <0.001 | 1472 (0.9) | <0.001 | 5991 (8.6) | <0.001 |
| Closed reduction with internal fixation | 59853 (16.7) | 36251 (24.7) | <0.001 | 24548 (15.5) | <0.001 | 2110 (3.0) | <0.001 |
| Open reduction without internal fixation | 4301 (1.2) | 1326 (0.9) | <0.001 | 2856 (1.8) | <0.001 | 502 (0.7) | <0.001 |
| Open reduction with internal fixation | 168809 (47.1) | 68313 (46.6) | <0.001 | 79299 (50.0) | <0.001 | 29658 (42.5) | <0.001 |
| Debridement | 60212 (16.8) | 13917 (9.5) | <0.001 | 37366 (23.5) | <0.001 | 5520 (7.9) | <0.001 |
| Unspecified bone surgery | 71 (0.02) | 57 (0.1) | <0.001 | 39 (0.02) | <0.001 | 7 (0.01) | <0.001 |
| Complications, n (%) | |||||||
| Amputation through bone | 5017 (1.4) | 1862 (1.3) | <0.001 | 2436 (1.5) | <0.001 | 594 (0.9) | <0.001 |
| Acute kidney injury | 3916 (1.1) | 1954 (1.3) | <0.001 | 1486 (0.9) | <0.001 | 918 (1.3) | <0.001 |
| ARDS | 6039 (1.7) | 3262 (2.2) | <0.001 | 2238 (1.4) | <0.001 | 1421 (2.0) | <0.001 |
| Myocardial infarction | 1044 (0.3) | 504 (0.3) | <0.001 | 376 (0.2) | <0.001 | 240 (0.3) | <0.05 |
| Pulmonary embolism | 3188 (0.9) | 1917 (1.3) | <0.001 | 1165 (0.7) | <0.001 | 432 (0.6) | <0.001 |
| Unplanned ICU | 1742 (0.5) | 802 (0.5) | <0.001 | 768 (0.5) | 0.86 | 349 (0.5) | 0.57 |
| Unplanned intubation | 2468 (0.7) | 1152 (0.8) | <0.05 | 1057 (0.7) | 0.14 | 537 (0.8) | <0.05 |
| Osteomyelitis | 177 (0.05) | 59 (0.01) | <0.05 | 115 (0.1) | <0.001 | 24 (0.03) | <0.05 |
| Pneumonia | 11975 (3.3) | 5785 (3.9) | <0.001 | 4783 (3.0) | <0.001 | 2969 (4.3) | <0.001 |
| Mortality, n (%) | 16152 (4.5) | 8889 (6.2) | <0.001 | 4473 (2.9) | <0.001 | 4806 (7.1) | <0.001 |
SD = standard deviation, IQR = interquartile range, LOS = length of stay, ICU = intensive care unit, ARDS = acute respiratory distress syndrome.
3.2. Demographics of patients with acute osteomyelitis
Compared to those without acute osteomyelitis, trauma patients with acute osteomyelitis were younger (mean age, 47.8 vs. 49.6, p < 0.001), less likely to be involved in a blunt mechanism (79.6% vs. 95.4%, p < 0.001) and had a higher median ISS (10.0 vs. 8.0, p < 0.001). Patients with acute osteomyelitis had a higher rate of associated spine injury (28.7% vs. 16.4%, p < 0.001) and isolated lower extremity fracture (38.2% vs. 19.8%, p < 0.001), but less TBI (22.4% vs. 30.7%, p < 0.001) and isolated upper extremity fracture (11.0% vs. 14.5%, p < 0.05) (Table 4).
Table 4.
Patient characteristics of all adult trauma patients with and without osteomyelitis.
| Characteristic | - Osteomyelitis |
+ Osteomyelitis |
p-value |
|---|---|---|---|
| (n = 4788970) | (n = 833) | ||
| Age, years, mean (SD) | 49.6 (21) | 47.8 (17) | <0.001 |
| Sex, male, n (%) | 3489253 (63.7) | 597 (71.7) | <0.001 |
| Blunt mechanism, n (%) | 4570498 (95.4) | 663 (79.6) | <0.001 |
| ISS, median (IQR) | 8.0 (9) | 10.0 (7) | <0.001 |
| Injuries, n (%) | |||
| Traumatic brain injury | 1687623 (30.7) | 187 (22.4) | <0.001 |
| Spine | 900861 (16.4) | 239 (28.7) | <0.001 |
| Isolated upper extremity fracture | 796072 (14.5) | 92 (11.0) | <0.05 |
| Isolated lower extremity fracture | 1088474 (19.8) | 318 (38.2) | <0.001 |
| Upper and lower extremity fracture | 196676 (3.6) | 135 (16.2) | <0.001 |
SD = standard deviation, ISS = injury severity score, IQR = interquartile range.
3.3. Univariable analysis for risk of acute osteomyelitis
On univariable analysis for risk of osteomyelitis in adult trauma patients, the strongest risk factors, in order, were non-pseudomonas bacteremia (OR 41.85, 95% CI 28.49–61.48, p < 0.001), PAD (OR 11.14, 95% CI 7.59–16.35, p < 0.001) and open tibia/fibula shaft fracture (OR 9.30, 95% CI 7.26–11.92, p < 0.001). Pseudomonas bacteremia and long-term steroid use were not associated with an increased risk for acute osteomyelitis (p > 0.05) (Table 5).
Table 5.
Univariable analysis of risk factors for osteomyelitis in adult trauma patients.
| Risk factor | OR | CI | p value |
|---|---|---|---|
| Closed femur shaft fracture | 2.07 | 1.49–2.88 | <0.001 |
| Open femur shaft fracture | 5.69 | 3.76–8.62 | <0.001 |
| Femur major surgery | 2.27 | 1.88–2.75 | <0.001 |
| Closed tibia/fibula shaft fracture | 3.59 | 2.74–4.70 | <0.001 |
| Open tibia/fibula shaft fracture | 9.30 | 7.26–11.92 | <0.001 |
| Tibia/fibula major surgery | 5.77 | 4.99–6.67 | <0.001 |
| Closed humerus shaft fracture | 1.38 | 0.78–2.45 | 0.27 |
| Open humerus shaft fracture | 6.92 | 4.00–11.97 | <0.001 |
| Humerus major surgery | 3.12 | 2.31–4.20 | <0.001 |
| Age ≥65 | 0.53 | 0.44–0.63 | <0.001 |
| Steroid use | 1.19 | 0.44–3.17 | 0.73 |
| Positive BAC on admission | 1.26 | 1.03–1.53 | <0.05 |
| Positive illegal drug screen on admission | 1.93 | 1.49–2.49 | <0.001 |
| Hypotension on admission | 2.82 | 2.20–3.61 | <0.001 |
| Smoker | 2.19 | 1.87–2.56 | <0.001 |
| Diabetes | 2.45 | 2.08–2.88 | <0.001 |
| ESRD | 3.17 | 2.03–4.94 | <0.001 |
| ISS≥25 | 3.46 | 2.90–4.14 | <0.001 |
| Blood transfusion | 5.56 | 4.79–6.46 | <0.001 |
| Peripheral arterial disease | 11.14 | 7.59–16.35 | <0.001 |
| Pseudomonas bacteremia | 5.45 | 0.27–8.59 | 0.98 |
| Non-pseudomonas bacteremia | 41.85 | 28.49–61.48 | <0.001 |
BAC = blood alcohol concentration, ESRD = end-stage renal disease, ISS = injury severity score.
3.4. Multivariable analysis for risk of acute osteomyelitis
After adjusting for covariates in a multivariable logistic regression analysis, non-pseudomonas bacteremia continued to be the strongest independent risk factor for osteomyelitis (OR 9.30, 95% CI 3.74–23.10, p < 0.001) followed by tibia/fibula major surgery (OR 5.31, 95% CI 3.81–7.41, p < 0.001), blood transfusion (OR 3.74, 95% CI 2.74–5.10, p < 0.001) and humerus major surgery (OR 2.74, 95% CI 1.57–4.77, p < 0.001). None of the long bone shaft fractures analyzed were found to be independent risk factors for acute osteomyelitis (p > 0.05). This remained true for both open and closed injuries (Table 6).
Table 6.
Multivariable analysisa for risk of osteomyelitis in adult trauma patients.
| Risk factor | OR | CI | p value |
|---|---|---|---|
| Closed femur shaft fracture | 0.82 | 0.43–1.58 | 0.55 |
| Open femur shaft fracture | 1.54 | 0.75–3.19 | 0.24 |
| Femur major surgery | 1.37 | 0.85–2.22 | 0.19 |
| Closed tibia/fibula shaft fracture | 1.28 | 0.79–2.07 | 0.32 |
| Open tibia/fibula shaft fracture | 1.21 | 0.71–2.04 | 0.49 |
| Tibia/fibula major surgery | 5.31 | 3.81–7.41 | <0.001 |
| Closed humerus shaft fracture | 0.18 | 0.04–0.79 | <0.05 |
| Open humerus shaft fracture | 2.03 | 0.83–4.97 | 0.12 |
| Humerus major surgery | 2.74 | 1.57–4.77 | <0.001 |
| Age ≥65 | 0.81 | 0.51–1.29 | 0.37 |
| Steroid use | 0.98 | 0.77–1.22 | 0.45 |
| Positive BAC on admission | 0.92 | 0.69–1.23 | 0.55 |
| Positive illegal drug screen on admission | 1.57 | 1.20–2.08 | <0.05 |
| Smoker | 1.94 | 1.44–2.62 | <0.001 |
| Hypotension on admission | 2.08 | 1.37–3.14 | <0.05 |
| Diabetes | 2.10 | 1.42–3.12 | <0.001 |
| ESRD | 3.37 | 1.04–10.90 | <0.05 |
| ISS≥25 | 2.29 | 1.67–3.14 | <0.001 |
| Blood transfusion | 3.74 | 2.74–5.10 | <0.001 |
| Peripheral arterial disease | 1.88 | 0.26–13.81 | 0.54 |
| Pseudomonas bacteremia | 3.45 | 0.34–8.59 | 0.98 |
| Non-pseudomonas bacteremia | 9.30 | 3.74–23.10 | <0.001 |
BAC = blood alcohol concentration, ESRD = end-stage renal disease, ISS = injury severity score.
= controlled for diabetes, peripheral arterial disease, steroid use, smoker, positive blood alcohol concentration on admission, positive illegal drug screen on admission, age ≥65, end-stage renal disease, blood transfusion, bacteremia, hypotension on admission, major bone surgery and injury severity score ≥25.
4. Discussion
Our study examines a large national database of trauma patients with long bone shaft fractures and provides an analysis for the risk of acute osteomyelitis in trauma. The rate of long bone shaft fractures is 6.5% with most involving the tibia/fibula. The rate of acute osteomyelitis across all patients with long bone shaft fractures is less than 0.1%. The strongest independent risk factor for acute osteomyelitis in trauma patients is non-pseudomonas bacteremia. We did not find any long bone (humerus, femur, or tibia/fibula) shaft fracture to increase risk for acute osteomyelitis and this remained true for both open and closed injuries. However, major surgery of the tibia/fibula is associated with over a five-fold increased risk for acute osteomyelitis and major surgery of the humerus is associated with nearly a three-fold increased risk for acute osteomyelitis.
Contrary to our hypothesis, we did not find open or closed long bone shaft fractures to be associated with higher risk for acute osteomyelitis. This association has widely been believed to be true by practicing surgeons since the 1960s, even though the evidence has been poor.18,19 The pathogenesis of osteomyelitis is likely multifactorial but remains poorly understood. In trauma, it is believed to involve direct inoculation of microorganisms into bone made possible by breakage in skin and/or soft tissue.4 This is the basis of why open fractures may have higher risk for acute osteomyelitis compared to closed fractures. Only 18 cases of osteomyelitis following any closed bone fracture have been reported in the literature since the 1970s.20 However, we did not find open (or closed) long bone shaft fractures to be associated with higher risk for acute osteomyelitis in adult trauma patients. One reason for this may be that acute development of osteomyelitis is more related to hematogenous inoculation then direct spread from the wound, evidenced by our finding that non-pseudomonas bacteremia has the strongest association with acute development of osteomyelitis in the hospitalized trauma patient. Additional prospective studies to confirm the lack of correlation between long bone shaft fractures and acute osteomyelitis in adults is needed.
Bacteremia has been reported to be associated with the development of acute osteomyelitis.21,22 Jorge et al. found Pseudomonas infection to be associated with nearly a three-fold increased risk for post-traumatic osteomyelitis.17 However, the most commonly isolated organism in patients with osteomyelitis is reported to be Staphylococcus.23,24 In our study, Staphylococcus occurred in nearly 20% of patients and was the most common specified organism. We did not find Pseudomonas bacteremia to be associated with increased risk for acute osteomyelitis in trauma patients. In contrast, non-pseudomonas bacteremia was the strongest predictor of acute osteomyelitis. Future prospective studies that include intraoperative cultures, as well as the source, organism type and associated sensitivities of the organisms causing bacteremia-associated osteomyelitis appears warranted.
The decision to operate and the timing of the procedure may affect the rate of acute osteomyelitis in patients with long bone fractures. Although the classic teaching has been to debride devitalized and contaminated tissue within six hours of patients arriving with traumatic long bone fractures, previous studies have demonstrated that this is accomplished less than 50% of the time.7,25 Regardless, debridement was the second most common major bone surgery performed in all patients with long bone shaft fractures in our study. Hull et al. studied nearly 500 patients with open extremity fractures, excluding open hand fractures, and found that timing of debridement does not correlate with the rate of deep infections in patients with Gustilo-Anderson grade I open fractures, but did increase the rate of deep infection in patients with grade II or III injuries.26 However, a large systematic review of nearly 3600 open long bone fractures found no correlation with delayed operative debridement and rates of deep infection.2 It is important to note that the “six-hour rule” was coined before the era of modern antibiotics, which may play a more important role than the timing of surgical intervention.
Aside from the timing, surgery itself may be a risk factor for the development of post-operative infection. Merritt et al. studied trauma patients with open fractures undergoing operative intervention with or without fixation devices.8 Bacterial counts taken of debrided tissue in the beginning of the case did not significantly correlate with the development of post-operative infection while bacterial counts taken at the end of the case had a strong correlation. Furthermore, patients receiving external or internal fixation devices had a higher rate of post-operative infection compared to those with no implants. This may be due to concomitant soft tissue injury and colonization of implants. In contrast, Patzakis et al. studied over 1100 open extremity fractures and did not find a correlation between primary internal fixation and subsequent infection.10 Our study suggests that the correlation between major bone surgery and infection in patients with long bone shaft fractures may be location specific. We did not find major bone surgery on the femur to increase risk of acute osteomyelitis but operating on the tibia/fibula or humerus did increase the risk by five-fold and three-fold, respectively. This may be a function of the increased soft tissue coverage preventing direct inoculation of the bone as well as the excellent blood supply to the shaft of the femur. Compared to the femur, the tibia/fibula does not have the extra layers of subcutaneous tissue and musculature protecting it from trauma, particularly on the anteromedial face, where open injuries are most common. This may result in considerable associated internal soft tissue loss that may not be obvious on the initial survey. Early fixation of these injuries that appear relatively benign externally, may result in florid infection.27 Additional prospective studies to confirm the association between location-specific long bone shaft fracture, as well as if the timing of intervention correlates with rates of postoperative acute osteomyelitis, is warranted.
Perioperative blood transfusion may be associated with a higher rate of post-operative infection. Pulido et al. demonstrated that allogenic blood transfusion is associated with more than a two-fold increased risk for periprosthetic joint infection in patients undergoing hip or knee arthroplasty.28 This association was subsequently confirmed in a large meta-analysis of patients undergoing total joint arthroplasty.29 We found perioperative blood transfusion to also be associated with increased risk of acute osteomyelitis in trauma patients. However, this may simply be a surrogate for severe injuries, surrounding soft tissue damage and perioperative hemodynamic instability, which may invalidate this finding. Future prospective studies examining the role that intraoperative hemodynamic instability and grade of the extremity fracture have in the development of post-traumatic osteomyelitis is warranted.
As a retrospective database study, our study has several limitations. All data fields in the NTDB are subject to input error. In addition, the difficulty in diagnosing osteomyelitis and the lack of a uniform definition for osteomyelitis used by all participating centers in the NTDB certainly introduced variability in reporting. Furthermore, the NTDB does not track patient outcomes after discharge. As osteomyelitis can develop over several days to weeks, some patients may have developed osteomyelitis after discharge, but were not included in the study. Fields relevant to our study that are missing in the NTDB include units of blood transfusion received, Gustilo-Anderson grade of open fractures, intraoperative hemodynamics, and tissue culture with quantification results. Additionally, many patients had unspecified bacteremia as the NTDB does not require documentation of the organism causing the infection. Thus, in our analysis of pseudomonas versus non-pseudomonas bacteremia, some cases of pseudomonas bacteremia may not have been reported as such and been included in the non-pseudomonas group.
5. Conclusion
In adult patients, long bone shaft fracture is not independently associated with increased risk for acute osteomyelitis. Major extremity surgery on the humerus and tibia/fibula, but not femur, are risk factors for osteomyelitis. Additionally, perioperative blood transfusion increases risk for acute osteomyelitis in trauma. However, the most significant risk factor for development of acute osteomyelitis is non-pseudomonas bacteremia.
Conflict of interest statement
The authors declare that they have no conflict of interest.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.jcot.2019.04.003.
Appendix A. ICD-9 diagnosis codes for long bone shaft fractures
| Femur | Tibia/fibula | Humerus |
|---|---|---|
| 821–821.01 (closed) | 823.2–823.22 (closed) | 812.2–812.21 (closed) |
| 821.1–821.11 (open) | 823.3–823.32 (open) | 812.3–812.31 (open) |
Appendix A. Supplementary data
The following is the Supplementary data to this article:
References
- 1.Klenerman L. A history of osteomyelitis from the journal of bone and joint surgery. J Bone Jt Surg. 2007;89(5):667–670. doi: 10.1302/0301-620X.89B5.19170. [DOI] [PubMed] [Google Scholar]
- 2.Schenker M.L., Yannascoli S., Baldwin K.D., Ahn J., Mehta S. Does Timing to operative debridement affect infectious complications in open long-bone fractures?: a systematic review. J Bone Jt Surg. 2012;94(12):1057–1064. doi: 10.2106/JBJS.K.00582. [DOI] [PubMed] [Google Scholar]
- 3.Waldvogel F.A., Medoff G., Swartz M.N. Osteomyelitis: a review of clinical features, therapeutic considerations and unusual aspects. N Engl J Med. 1970;282(4):198–206. doi: 10.1056/NEJM197001222820406. [DOI] [PubMed] [Google Scholar]
- 4.Calhoun J.H., Manring M., Shirtliff M., editors. Seminars in Plastic Surgery. © Thieme Medical Publishers; 2009. Osteomyelitis of the long bones. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Court-Brown C., Rimmer S., Prakash U., McQueen M. The epidemiology of open long bone fractures. Injury. 1998;29(7):529–534. doi: 10.1016/s0020-1383(98)00125-9. [DOI] [PubMed] [Google Scholar]
- 6.Trampuz A., Widmer A.F. Infections associated with orthopedic implants. Curr Opin Infect Dis. 2006;19(4):349–356. doi: 10.1097/01.qco.0000235161.85925.e8. [DOI] [PubMed] [Google Scholar]
- 7.Robson M.C., Duke W.F., Krizek T.J. Rapid bacterial screening in the treatment of civilian wounds. J Surg Res. 1973;14(5):426–430. doi: 10.1016/0022-4804(73)90049-8. [DOI] [PubMed] [Google Scholar]
- 8.Merritt K. Factors increasing the risk of infection in patients with open fractures. J. Trauma Acute Care Surg. 1988;28(6):823–827. doi: 10.1097/00005373-198806000-00018. [DOI] [PubMed] [Google Scholar]
- 9.DeLong W.G., Born C.T., Wei S.Y., Petrik M.E., Ponzio R., Schwab C.W. Aggressive treatment of 119 open fracture wounds. J. Trauma Acute Care Surg. 1999;46(6):1049–1054. doi: 10.1097/00005373-199906000-00012. [DOI] [PubMed] [Google Scholar]
- 10.Patzakis M.J., Wilkins J. Factors influencing infection rate in open fracture wounds. Clin Orthop Relat Res. 1989;243:36–40. [PubMed] [Google Scholar]
- 11.Surgeons ACo . American College of Surgeons; 2016 2016. Committee onTrauma. American College of surgeons. NTDB, version 10. [Google Scholar]
- 12.Lavery L.A., Armstrong D.G., Wunderlich R.P., Mohler M.J., Wendel C.S., Lipsky B.A. Risk factors for foot infections in individuals with diabetes. Diabetes Care. 2006 Jun;29(6):1288–1293. doi: 10.2337/dc05-2425. PubMed PMID: 16732010. Epub 2006/05/30. eng. [DOI] [PubMed] [Google Scholar]
- 13.Abdelfadil E., Salem A.S., Mourad S.I., Al-Belasy F.A. Infected mandibular fractures: risk factors and management. J. Oral Hygiene Health. 2013:1–8. [Google Scholar]
- 14.Matos M.A., Lima L.G., de Oliveira L.A. Predisposing factors for early infection in patients with open fractures and proposal for a risk score. J Orthop Traumatol : Off. J. Italian Soc. Orthop. Traumatol. 2015 Sep;16(3):195–201. doi: 10.1007/s10195-015-0345-z. PubMed PMID: 25720585. Pubmed Central PMCID: PMC4559547. Epub 2015/02/28. eng. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Tribble D., Potter B., Lewandowski L., editors. Osteomyelitis Risk Factors Related to Combat Trauma Open Tibia Fracture. Open Forum Infectious Diseases. Oxford University Press; 2015. [Google Scholar]
- 16.Dounousi E., Duni A., Xiromeriti S., Pappas C., Siamopoulos K.C. Acute bacterial sternoclavicular osteomyelitis in a long-term renal transplant recipient. World J Transplant. 2016 Jun 24;6(2):442–446. doi: 10.5500/wjt.v6.i2.442. PubMed PMID: 27358791. Pubmed Central PMCID: PMC4919750. Epub 2016/07/01. eng. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Jorge L.S., Chueire A.G., Fucuta P.S. Predisposing factors for recurrence of chronic posttraumatic osteomyelitis: a retrospective observational cohort study from a tertiary referral center in Brazil. Patient Saf Surg. 2017;11:17. doi: 10.1186/s13037-017-0133-1. PubMed PMID: 28588645. Pubmed Central PMCID: PMC5457537. Epub 2017/06/08. eng. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rowe C.R., Sakellarides H.T. Recent advances in treatment of osteomyelitis following fractures of the long bones. Surgical Clinics of North America. 1961;41(6):1593–1605. doi: 10.1016/s0039-6109(16)36531-8. PubMed PMID 14494633. [DOI] [PubMed] [Google Scholar]
- 19.Errico M.J. INCIDENCE OF OSTEOMYELITIS IN NONUNION OF FRACTURE OF THE SHAFT OF THE TIBIA. J. Trauma Acute Care Surg. 1967;7(6):838–843. doi: 10.1097/00005373-196711000-00005. [DOI] [PubMed] [Google Scholar]
- 20.Kocutar T., Snoj Z., Salapura V. Complicated acute haematogenous osteomyelitis with fatal outcome following a closed clavicle fracture-a case report and literature review. Bjr Case Reports. 2016;2(2) doi: 10.1259/bjrcr.20150389. 20150389. PubMed PMID: WOS:000406879600032. English. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ortwine J.K., Bhavan K. Morbidity, mortality, and management of methicillin-resistant S. aureus bacteremia in the USA: update on antibacterial choices and understanding. Hospital practice (1995) 2018 Feb 12:1–9. doi: 10.1080/21548331.2018.1435128. PubMed PMID: 29400119. Epub 2018/02/06. eng. [DOI] [PubMed] [Google Scholar]
- 22.Murillo O., Grau I., Lora-Tamayo J. The changing epidemiology of bacteraemic osteoarticular infections in the early 21st century. Clin. Microbiol. Infect. : Off. Publ. Eur. Soc. Clin. Microbiol. Infect. Dis. 2015 Mar;21(3):254.e1–254.e8. doi: 10.1016/j.cmi.2014.09.007. PubMed PMID: 25618436. Epub 2015/01/27. eng. [DOI] [PubMed] [Google Scholar]
- 23.Hadjipavlou A.G., Mader J.T., Necessary J.T., Muffoletto A.J. Hematogenous pyogenic spinal infections and their surgical management. Spine. 2000;25(13):1668–1679. doi: 10.1097/00007632-200007010-00010. [DOI] [PubMed] [Google Scholar]
- 24.Kumar A., Sandoe J., Kumar N. Three cases of vertebral osteomyelitis caused by Streptococcus dysgalactiae subsp. equisimilis. J Med Microbiol. 2005;54(11):1103–1105. doi: 10.1099/jmm.0.46061-0. [DOI] [PubMed] [Google Scholar]
- 25.Namdari S., Baldwin K.D., Matuszewski P., Esterhai J.L., Mehta S. Delay in surgical débridement of open tibia fractures: an analysis of national practice trends. J Orthop Trauma. 2011;25(3):140–144. doi: 10.1097/BOT.0b013e3181e3dff1. [DOI] [PubMed] [Google Scholar]
- 26.Hull P., Johnson S., Stephen D., Kreder H., Jenkinson R. Delayed debridement of severe open fractures is associated with a higher rate of deep infection. Bone Joint J. 2014;96(3):379–384. doi: 10.1302/0301-620X.96B3.32380. [DOI] [PubMed] [Google Scholar]
- 27.Chaudhary M.M. Infected nonunion of tibia. Indian J Orthop. 2017;51(3):256. doi: 10.4103/ortho.IJOrtho_199_16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Pulido L., Ghanem E., Joshi A., Purtill J.J., Parvizi J. Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop Relat Res. 2008;466(7):1710–1715. doi: 10.1007/s11999-008-0209-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zhu Y., Zhang F., Chen W., Liu S., Zhang Q., Zhang Y. Risk factors for periprosthetic joint infection after total joint arthroplasty: a systematic review and meta-analysis. J Hosp Infect. 2015;89(2):82–89. doi: 10.1016/j.jhin.2014.10.008. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.