Translate this page into:
Preoperative blood transfusion is an independent risk factor for postoperative infection, readmission, and mortality following surgery for femoral shaft fractures
, Bruce B. Zhang, David H. Mai, Carl B. Paulino, Qais Naziri, Aden N. Malik
⁎Corresponding author: Jason M. Dayan. Jason.dayan@downstate.edu
-
Received: ,
Accepted: ,
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
Abstract
Abstract
Femoral shaft fractures (FSFs) are often associated with significant blood loss, frequently necessitating preoperative blood transfusion, which may increase the risk of complications. This study evaluated the relationship between preoperative transfusion and postoperative outcomes, including infection, readmission, reoperation, hemodynamic complications, and mortality in patients undergoing surgical fixation for FSF.
A retrospective cohort analysis was conducted using the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database from 2012 to 2021. Adult patients (≥18 years) who underwent surgical fixation for FSF were included. The primary exposure was transfusion of ≥1 unit of whole or packed red blood cells within 72 h before surgery. Patients were grouped by transfusion status. Demographic, clinical, and procedural variables were compared using chi-square tests, and multivariable regression adjusted for potential confounders. Primary outcomes were 30-day postoperative infection, hospital readmission, reoperation, hemodynamic complications, and mortality.
Of 6812 patients identified, 373 (5.5 %) received a preoperative transfusion. Most patients were 80–89 years, White, female, non-Hispanic, ASA 3, with normal BMI, independent functional status, hypertension, low hematocrit, normal WBC, and non-emergent cases. Most were nonsmokers and without major comorbidities (COPD, CHF, cancer, diabetes, immunosuppression, bleeding disorders, or weight loss) [Table 1]. After adjustment, preoperative transfusion was independently associated with increased odds of postoperative infection (OR 1.67, 95 % CI 1.12–2.42, p = 0.009), hospital readmission (OR 1.49, 95 % CI 1.03–2.13, p = 0.031), and mortality (OR 1.73, 95 % CI 1.15–2.55, p = 0.007). No significant associations were found for reoperation (p = 0.207) or hemodynamic complications (p = 0.57) [Table 4].
Preoperative blood transfusion in patients undergoing surgical fixation for femoral shaft fractures is associated with higher risks of infection, readmission, and mortality. No significant differences were observed for hemodynamic complications or reoperation. These findings can help guide surgeons as they balance optimal hemoglobin levels preoperatively with risks of postoperative complications associated with transfusion.
Level III.
Keywords
Femoral shaft fracture
preoperative transfusion
Infection
Urinary tract infection
Mortality
Readmission
1 Introduction
Femoral shaft fractures (FSF) have a global incidence between 10 and 45.5 in 100,000 per year1–3 and a 5-year mortality risk of 58.5 %.4 Common causes of FSF include high-energy mechanisms in younger individuals, as well as low-energy trauma in older individuals with osteoporotic bone.5 Less common causes of FSF include pathologic fractures, stress fractures from overuse in athletes and military recruits, and atypical fractures associated with bisphosphonate use.1,6,7 Definitive surgical fixation is most often achieved with intramedullary nailing (IMN).
A common complication associated with FSF is acute blood loss, ranging between 1 and 3 L, which can result in anemia and hemorrhagic shock.8,9 Excessive blood loss has been shown to be a negative prognostic factor in recovery from orthopedic surgeries, and is associated with increased postoperative morbidity, mortality, and hospital length of stay, as well as poorer functional recovery.10–12 Therefore, blood transfusions are used in perioperative blood management in up to a fourth of patients with FSF.13 While blood transfusions can be lifesaving, they are not without risks. They have been shown to be associated with increased risk of infection, immune dysregulation, morbidity, and post-operative hospital length of stay.14–20 While prior studies acknowledge the increased risk of blood loss in FSF, to the best of our knowledge, no study has assessed the association between adverse postoperative events following surgical fixation for FSF and preoperative blood transfusions. Thus, the primary goal of this study is to examine potential postoperative complications of patients who received and did not receive preoperative transfusions among patients who underwent surgical fixation for FSF. We hypothesize that preoperative transfusion will be associated with higher rates of adverse postoperative outcomes, including infection, readmission, reoperation, hemodynamic complications, and mortality. A secondary goal of this study is to determine risk factors for preoperative blood transfusion among patients who underwent surgical fixation for FSF. We hypothesize that older, sicker patients and patients with bleeding disorders will have a higher rate of preoperative transfusion.
2 Materials and methods
2.1 Data collection and patient selection
A retrospective cohort study was performed using the American College of Surgeons – National Surgical Quality Improvement Program (ACS – NSQIP) database. The current procedural terminology (CPT) codes 27506 and 27507 were used to identify all patients who underwent surgical treatment for femoral shaft fracture between the years 2012 and 2021. These CPT codes represent any surgical treatment for femoral shaft fractures, including internal fixation (screws, plates, intramedullary nails, or other fixation devices). Patients under the age of 18 were excluded from the study. The remaining patients were split into two cohorts: those who received a preoperative transfusion and those who did not. Preoperative transfusion is defined as ≥1 unit of whole or packed red blood cells in the 72-h period prior to surgery. A total of 6980 patients were identified, of which 6607 (94.66 %) did not receive a preoperative transfusion and 373 (5.34 %) did receive a preoperative transfusion.
2.2 Variables
Patient characteristics included basic demographics, pre-operative health status, and procedural characteristics. Basic demographics included age, race (White, Black, other, unknown), ethnicity (Hispanic, Non-Hispanic, unknown), and sex. Pre-operative health status included the following: body mass index (BMI), chronic immunosuppression use, chronic obstructive pulmonary disease (COPD), chronic heart failure (CHF), disseminated cancer, hypertension, bleeding disorders, diabetes mellitus (no, diabetes controlled with insulin, diabetes controlled with oral agents), smoking, preoperative functional status, preoperative hematocrit, and preoperative white blood cell (WBC) count. Procedural characteristics included American Society of Anesthesiologists (ASA) classification and emergent case designation. Preoperative hematocrit ranges were determined based on the reference ranges provided by the World Health Organization. Normal hematocrit ranges in adults were 41–50 % for men and 36–44 % for women.21,22 Preoperative WBC ranges were determined based on the reference ranges provided by the American Association for Clinical Chemistry. The normal WBC range in adults was 4000–11,000 cells/μL.23
2.3 Outcomes
The following 30-day outcomes were analyzed: infection, hospital readmission, reoperation, hemodynamic complications, and mortality. Infection included any cases of pneumonia, sepsis, and septic shock within 7 days, as well as superficial surgical site infection (SSI), deep SSI, periprosthetic joint infection (PJI), and urinary tract infection (UTI). Hemodynamic complications included occurrences of myocardial infarction (MI), pulmonary embolism (PE), deep vein thrombosis (DVT), stroke, and cardiac arrest.
2.4 Statistical analysis
R software, version 4.4.1 was used for all data analysis.24 Chi-squared and Fisher's exact tests (used when at least one cell count was less than 5) were used to evaluate categorical variables when analyzing variable-related differences between the transfused and non-transfused groups. All included variables were added to a multivariable logistic regression model, which was used to calculate the odds ratio of receiving preoperative blood transfusion. The variables were then included in another multivariable logistic regression model, which was used to calculate the odds ratio for each included variable of infection, readmission, reoperation, hemodynamic complications, and mortality within thirty days following the operation, along with a 95 % confidence interval (CI). All variables were categorical and were presented as counts (with percentages) within their groups. P-values <0.05 were considered significant.
2.5 Source of funding
This study had no external funding.
3 Results
Over the study period, there were 6812 cases of surgical fixation for femoral shaft fracture, among which 373 (5.48 %) received preoperative transfusion and 6439 (94.52 %) did not receive a preoperative transfusion. In both cohorts, the highest proportion of patients were aged 80–89 years old, White, non-Hispanic, and female. They had normal BMI, hypertension, independent preoperative functional status, low preoperative hematocrit levels, normal preoperative WBC counts, ASA classification of 3, and non-emergent case designation. Also, most patients in both cohorts did not smoke or have any of the following: chronic immunosuppression use, COPD, CHF, cancer, bleeding disorders, weight loss, or diabetes. [Table 1].
| No Preoperative Transfusion | Preoperative Transfusion | p-valuea | |||
| N = 6607 | N = 373 | ||||
| Basic Demographicsrowhead | |||||
| Age (years) | <0.001 | ||||
| 18–59 | 1583 (24 %) | 47 (13 %) | |||
| 60–69 | 1220 (18 %) | 71 (19 %) | |||
| 70–79 | 1492 (23 %) | 73(20 %) | |||
| 80–89 | 1533 (23 %) | 105 (28 %) | |||
| ≥90 | 779 (12 %) | 77 (21 %) | |||
| Race | 0.007 | ||||
| White | 4353 (66 %) | 253 (68 %) | |||
| Black/African American | 572 (8.7 %) | 47 (13 %) | |||
| Other | 337 (5.1 %) | 18 (4.8 %) | |||
| Unknown | 1345 (20 %) | 55 (15 %) | |||
| Ethnicity | 0.008 | ||||
| Not Hispanic | 4787 (73 %) | 277 (75 %0 | |||
| Hispanic | 568 (8.7 %) | 44 (12 %) | |||
| Unknown | 1211 (18 %) | 49 (13 %) | |||
| Sex | 0.216 | ||||
| Female | 4455 (67 %) | 266 (71 %) | |||
| Male | 2152 (33 %) | 107 (29 %) | |||
| Baseline Health Statusrowhead | |||||
| BMI | 0.011 | ||||
| Normal Weight (18.5 ≤ 25) | 1995 (33 %) | 126 (36 %) | |||
| Underweight (≤18.5) | 297 (4.9 %) | 26 (7.4 %) | |||
| Overweight (25 ≤ 30) | 1807 (30 %) | 80 (23 %) | |||
| Obese (≥30) | 1927 (32 %) | 121 (34 %) | |||
| Chronic Immunosuppression | 385 (5.8 %) | 37 (9.9 %) | 0.001 | ||
| COPD | 541 (8.2 %) | 33 (8.8 %) | 0.652 | ||
| Congestive Heart Failure | 261 (4.0 %) | 25 (6.7 %) | 0.009 | ||
| Disseminated Cancer | 509 (7.7 %) | 42 (11 %) | 0.013 | ||
| Hypertension | 3757 (57 %) | 263 (71 %) | <0.001 | ||
| Bleeding Disorders | 747 (11 %) | 88 (24 %) | <0.001 | ||
| Diabetes | <0.001 | ||||
| No Diabetes | 5273 (80 %) | 273 (73 %) | |||
| Treated with Insulin | 626 (9.5 %) | 63 (17 %) | |||
| Treated with Oral Agent | 708 (11 %) | 37 (9.9 %) | |||
| Smoking | 1063 (16 %) | 41 (11 %) | 0.009 | ||
| Preoperative Functional Status | <0.001 | ||||
| Independent | 5429 (83 %) | 252 (69 %) | |||
| Partially Dependent | 874 (13 %) | 79 (22 %) | |||
| Totally Dependent | 245 (3.7 %) | 35 (9.6 %) | |||
| Preoperative Hematocrit | <0.001 | ||||
| Normal | 2249 (34 %) | 16 (4.3 %) | |||
| Low Hematocrit | 4246 (64 %) | 357 (96 %) | |||
| High Hematocrit | 112 (1.7 %) | 0 (0 %) | |||
| Preoperative WBC | <0.001 | ||||
| Normal | 4380 (68 %) | 257 (69 %) | |||
| Low WBC | 121 (1.9 %) | 20 (5.4 %) | |||
| HighWBC | 1933 (30 %) | 95 (26 %) | |||
| Procedure Characteristicsrowhead | |||||
| ASA Classification | <0.001 | ||||
| <3 | 1797 (27 %) | 23 (6.2 %) | |||
| 3 | 3713 (56 %) | 204 (55 %) | |||
| >3 | 1084 (16 %) | 144 (39 %) | |||
| Emergency Case Designation | 1736 (26 %) | 93 (25 %) | 0.566 | ||
| No Preoperative Transfusion | Preoperative Transfusion | p-value | |
| N = 6607 | N = 373 | ||
| Complications (%) | |||
| Mortality | 3.8 | 11 | < 0.001 |
| Readmission | 5.9 | 11 | < 0.001 |
| Reoperation | 2.5 | 3.8 | 0.130 |
| Infection | 5.5 | 10 | < 0.001 |
| Pneumonia | 1.2 | 2.1 | 0.137 |
| Sepsis | 0.3 | 0.3 | >0.999 |
| Septic Shock | 0.1 | 0.5 | 0.096 |
| Superficial SSI | 0.6 | 1.1 | 0.289 |
| Deep SSI | 0.3 | 0.3 | >0.999 |
| PJI | 0.2 | 0.3 | 0.391 |
| UTI | 3.4 | 6.4 | 0.002 |
| Hemodynamic | 4.2 | 5.1 | 0.392 |
| Myocardial Infarction | 1.0 | 0.8 | >0.999 |
| Pulmonary Embolism | 1.0 | 0.3 | 0.263 |
| DVT | 1.4 | 1.6 | 0.730 |
| Stroke | 0.7 | 1.1 | 0.530 |
| Cardiac Arrest | 0.6 | 1.3 | 0.097 |
Multivariable regression analysis demonstrated that the following variables were risk factors for preoperative transfusion: age, black race (adjusted odds ratio [OR] 1.44, 95 % CI 1.00 to 2.04, p = 0.033), Hispanic ethnicity (OR 1.52, 95 % CI 1.04 to 2.17, p = 0.026), bleeding disorders (OR 1.58, 95 % CI 1.19 to 2.07, p = 0.001), preoperative functional status, low preoperative hematocrit levels (OR 9.88, 95 % CI 5.86 to 18.3, p < 0.001), low preoperative WBC count (OR 1.93, 95 % CI 1.08 to 3.27, p = 0.020), and ASA classification [Table 3].
| Preoperative Transfusion | |||
| OR | 95 % CI | P Value | |
| Basic Demographics | |||
| Age (years) | |||
| 18–59 | – | – | – |
| 60–69 | 1.25 | 0.83–1.90 | 0.297 |
| 70–79 | 1.09 | 0.72–1.66 | 0.693 |
| 80–89 | 1.24 | 0.82–1.89 | 0.318 |
| ≥90 | 1.64 | 1.04–2.60 | 0.033 |
| Race | |||
| White | – | – | – |
| Black/African American | 1.44 | 1.00–2.04 | 0.044 |
| Other | 1.17 | 0.65–1.95 | 0.581 |
| Unknown | 0.77 | 0.46–1.25 | 0.303 |
| Ethnicity | |||
| Not Hispanic | – | – | – |
| Hispanic | 1.52 | 1.04–2.17 | 0.026 |
| Unknown | 1.02 | 0.60–1.68 | 0.947 |
| Sex | |||
| Female | – | – | – |
| Male | 0.89 | 0.69–1.14 | 0.369 |
| Baseline Health Status | |||
| BMI | |||
| Normal Weight (18.5 ≤ 25) | – | – | – |
| Underweight (≤18.5) | 1.17 | 0.72–1.83 | 0.515 |
| Overweight (25 ≤ 30) | 0.76 | 0.56–1.03 | 0.081 |
| Obese (≥30) | 1.07 | 0.80–1.43 | 0.654 |
| Chronic Immunosuppression | 1.33 | 0.88–1.97 | 0.161 |
| COPD | 0.75 | 0.49–1.11 | 0.166 |
| Congestive Heart Failure | 0.94 | 0.58–1.47 | 0.798 |
| Disseminated Cancer | 1.02 | 0.68–1.49 | 0.923 |
| Hypertension | 1.11 | 0.85–1.46 | 0.440 |
| Bleeding Disorders | 1.58 | 1.19–2.07 | 0.001 |
| Diabetes | |||
| No Diabetes | – | – | – |
| Treated with Insulin | 1.18 | 0.84–1.64 | 0.323 |
| Treated with Oral Agent | 0.67 | 0.44–0.99 | 0.052 |
| Smoking | 0.86 | 0.58–1.25 | 0.448 |
| Preoperative Functional Status | |||
| Independent | – | – | – |
| Partially Dependent | 1.39 | 1.04–1.84 | 0.023 |
| Totally Dependent | 1.70 | 1.09–2.58 | 0.016 |
| Preoperative Hematocrit | |||
| Normal | – | – | – |
| Low Hematocrit | 9.88 | 5.86–18.3 | <0.001 |
| High Hematocrit | 0.00 | 0.00–0.00 | 0.974 |
| Preoperative WBC | |||
| Normal | – | – | – |
| Low WBC | 1.93 | 1.08–3.27 | 0.020 |
| HighWBC | 0.95 | 0.73–1.23 | 0.715 |
| Procedure Characteristics | |||
| ASA Classification | |||
| <3 | – | – | – |
| 3 | 2.93 | 1.81–5.05 | <0.001 |
| >3 | 5.86 | 3.49–10.3 | <0.001 |
| Emergency Case Designation | 0.98 | 0.74–1.29 | 0.904 |
Postoperative thirty-day complication rates are listed in Table II. Univariate analysis identified significantly higher rates of overall infection (10 % versus 5.5 %, p < 0.001), UTI (6.4 % versus 3.4 %, p = 0.002), readmission (11 % versus 5.9 %, p < 0.001), and mortality (11 % versus 3.8 %, p < 0.001) in the transfused group compared to the non-transfused group. [Table 2].
3.1 Infection
Compared to patients who did not receive preoperative transfusion, those who received preoperative transfusion had 2.01 (95 % CI 1.40 to 2.82, p < 0.001) times higher crude odds of infection. After adjusting for confounding variables, multivariable regression analysis demonstrated 1.67 (95 % CI 1.12 to 2.42, p = 0.009) higher odds of infection in patients who received a preoperative transfusion [Table 4].
| Mortality | Readmission | Reoperation | Infection | Hemodynamic Complications | |||||||||||
| OR | 95 % CI | P Value | OR | 95 % CI | P Value | OR | 95 % CI | P Value | OR | 95 % CI | P Value | OR | 95 % CI | P Value | |
| Preop. Transfusion | 1.73 | 1.15–2.55 | 0.007 | 1.49 | 1.03–2.13 | 0.031 | 1.47 | 0.77–2.60 | 0.207 | 1.67 | 1.12–2.42 | 0.009 | 0.86 | 0.49–1.41 | 0.571 |
| Basic Demographics | |||||||||||||||
| Age (years) | |||||||||||||||
| 18–59 | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| 60–69 | 1.27 | 0.72–2.30 | 0.413 | 1.22 | 0.84–1.79 | 0.301 | 1.81 | 1.07–3.09 | 0.030 | 1.12 | 0.73–1.73 | 0.593 | 1.94 | 1.22–3.14 | 0.006 |
| 70–79 | 1.37 | 0.78–2.47 | 0.287 | 1.64 | 1.14–2.37 | 0.008 | 1.68 | 0.99–2.91 | 0.060 | 1.16 | 0.77–1.77 | 0.488 | 1.66 | 1.03–2.71 | 0.040 |
| 80–89 | 2.83 | 1.66–5.02 | <0.001 | 1.52 | 1.03–2.24 | 0.035 | 1.38 | 0.77–2.50 | 0.289 | 1.81 | 1.20–2.75 | 0.005 | 2.20 | 1.36–3.62 | 0.002 |
| ≥90 | 6.36 | 3.63–11.5 | <0.001 | 1.72 | 1.10–2.71 | 0.018 | 0.93 | 0.42–1.97 | 0.849 | 1.63 | 1.01–2.64 | 0.045 | 2.20 | 1.26–3.88 | 0.006 |
| Race | |||||||||||||||
| White | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Black/African American | 1.13 | 0.70–1.76 | 0.596 | 1.43 | 1.02–1.99 | 0.034 | 0.87 | 0.47–1.51 | 0.645 | 1.16 | 0.77–1.69 | 0.466 | 1.88 | 1.26–2.73 | 0.001 |
| Other | 0.77 | 0.35–1.49 | 0.472 | 0.88 | 0.50–1.44 | 0.621 | 1.39 | 0.69–2.53 | 0.320 | 0.62 | 0.30–1.13 | 0.148 | 0.71 | 0.31–1.37 | 0.348 |
| Unknown | 0.76 | 0.40–1.41 | 0.394 | 0.60 | 0.36–0.97 | 0.044 | 0.77 | 0.39–1.48 | 0.447 | 1.06 | 0.68–1.64 | 0.784 | 0.77 | 0.45–1.30 | 0.338 |
| Ethnicity | |||||||||||||||
| Not Hispanic | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Hispanic | 0.43 | 0.21–0.79 | 0.011 | 1.04 | 0.70–1.50 | 0.840 | 0.65 | 0.32–1.21 | 0.211 | 0.71 | 0.44–1.10 | 0.146 | 1.14 | 0.71–1.77 | 0.562 |
| Unknown | 0.83 | 0.44–1.52 | 0.561 | 0.88 | 0.52–1.44 | 0.615 | 1.25 | 0.64–2.36 | 0.506 | 1.47 | 0.94–2.29 | 0.087 | 1.28 | 0.75–2.15 | 0.357 |
| Sex | |||||||||||||||
| Female | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Male | 1.28 | 0.95–1.71 | 0.101 | 1.02 | 0.81–1.30 | 0.842 | 1.02 | 0.71–1.46 | 0.909 | 0.86 | 0.66–1.12 | 0.278 | 1.05 | 0.79–1.39 | 0.731 |
| Baseline Health Status | |||||||||||||||
| BMI | |||||||||||||||
| Normal Weight (18.5 ≤ 25) | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Underweight (≤18.5) | 1.82 | 1.15–2.82 | 0.008 | 0.73 | 0.42–1.20 | 0.242 | 0.59 | 0.20–1.37 | 0.270 | 0.70 | 0.38–1.19 | 0.209 | 0.82 | 0.40–1.50 | 0.539 |
| Overweight (25 ≤ 30) | 0.79 | 0.56–1.10 | 0.169 | 0.76 | 0.58–1.01 | 0.057 | 0.92 | 0.60–1.39 | 0.683 | 0.95 | 0.72–1.26 | 0.748 | 0.97 | 0.70–1.35 | 0.875 |
| Obese (≥30) | 0.60 | 0.41–0.87 | 0.007 | 0.98 | 0.75–1.28 | 0.861 | 1.22 | 0.82–1.82 | 0.336 | 0.88 | 0.66–1.18 | 0.412 | 1.15 | 0.83–1.59 | 0.403 |
| Chronic Immunosuppression | 1.07 | 0.64–1.71 | 0.779 | 1.09 | 0.72–1.59 | 0.681 | 0.88 | 0.42–1.63 | 0.706 | 1.13 | 0.72–1.72 | 0.572 | 1.30 | 0.80–2.02 | 0.262 |
| COPD | 1.76 | 1.19–2.55 | 0.004 | 1.51 | 1.09–2.07 | 0.011 | 0.93 | 0.51–1.59 | 0.812 | 1.65 | 1.17–2.30 | 0.004 | 1.04 | 0.67–1.56 | 0.867 |
| Congestive Heart Failure | 1.50 | 0.93–2.55 | 0.083 | 1.62 | 1.07–2.39 | 0.019 | 0.98 | 0.40–2.01 | 0.952 | 1.04 | 0.63–1.64 | 0.874 | 1.74 | 1.07–2.72 | 0.019 |
| Disseminated Cancer | 4.47 | 3.05–6.50 | <0.001 | 2.57 | 1.88–3.46 | <0.001 | 1.38 | 0.79–2.28 | 0.225 | 1.13 | 0.73–1.69 | 0.560 | 1.83 | 1.21–2.68 | 0.003 |
| Hypertension | 0.91 | 0.67–1.24 | 0.526 | 1.01 | 0.79–1.29 | 0.964 | 1.14 | 0.79–1.66 | 0.493 | 1.45 | 1.11–1.92 | 0.008 | 1.23 | 0.91–1.68 | 0.181 |
| Bleeding Disorders | 1.81 | 1.31–2.47 | <0.001 | 1.21 | 0.90–1.60 | 0.189 | 0.98 | 0.59–1.55 | 0.929 | 1.20 | 0.88–1.62 | 0.239 | 1.51 | 1.08–2.08 | 0.014 |
| Diabetes | |||||||||||||||
| No Diabetes | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Treated with Insulin | 2.18 | 1.46–3.22 | <0.001 | 1.45 | 1.05–1.98 | 0.020 | 1.35 | 0.82–2.15 | 0.224 | 1.52 | 1.07–2.13 | 0.016 | 1.00 | 0.66–1.49 | 0.990 |
| Treated with Oral Agent | 0.89 | 0.54–1.41 | 0.632 | 1.12 | 0.80–1.55 | 0.490 | 0.94 | 0.54–1.54 | 0.811 | 1.23 | 0.86–1.73 | 0.234 | 0.94 | 0.62–1.38 | 0.757 |
| Smoking | 0.95 | 0.61–1.45 | 0.823 | 1.04 | 0.75–1.41 | 0.816 | 1.40 | 0.90–2.14 | 0.123 | 0.88 | 0.61–1.25 | 0.488 | 0.94 | 0.63–1.37 | 0.763 |
| Preoperative Functional Status | |||||||||||||||
| Independent | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Partially Dependent | 1.20 | 0.86–1.65 | 0.275 | 1.00 | 0.74–1.33 | 0.994 | 1.57 | 1.01–2.39 | 0.040 | 1.35 | 1.00–1.80 | 0.042 | 0.96 | 0.67–1.35 | 0.839 |
| Totally Dependent | 1.80 | 1.11–2.84 | 0.014 | 0.95 | 0.55–1.55 | 0.835 | 2.00 | 0.91–3.88 | 0.059 | 1.74 | 1.06–2.73 | 0.022 | 0.52 | 0.22–1.06 | 0.104 |
| Preoperative Hematocrit | |||||||||||||||
| Normal | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Low Hematocrit | 1.68 | 1.17–2.48 | 0.006 | 1.31 | 1.01–1.71 | 0.043 | 1.15 | 0.80–1.68 | 0.451 | 1.44 | 1.10–1.92 | 0.009 | 1.62 | 1.18–2.26 | 0.003 |
| High Hematocrit | 1.28 | 0.30–3.79 | 0.694 | 0.94 | 0.32–2.18 | 0.897 | 1.58 | 0.47–4.02 | 0.394 | 1.74 | 0.71–3.66 | 0.180 | 2.26 | 0.91–4.86 | 0.053 |
| Preoperative WBC | |||||||||||||||
| Normal | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Low WBC | 1.13 | 0.45–2.41 | 0.775 | 0.77 | 0.35–1.49 | 0.471 | 0.79 | 0.19–2.19 | 0.690 | 0.79 | 0.30–1.71 | 0.591 | 0.14 | 0.01–0.64 | 0.052 |
| HighWBC | 1.78 | 1.35–2.35 | <0.001 | 1.06 | 0.83–1.33 | 0.645 | 1.03 | 0.72–1.45 | 0.876 | 1.13 | 0.88–1.43 | 0.334 | 1.49 | 1.14–1.93 | 0.004 |
| Procedure Characteristics | |||||||||||||||
| ASA Classification | |||||||||||||||
| <3 | – | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| 3 | 4.26 | 2.09–10.2 | <0.001 | 1.86 | 1.31–2.71 | <0.001 | 1.04 | 0.68–1.63 | 0.858 | 1.06 | 0.76–1.49 | 0.739 | 1.40 | 0.94–2.12 | 0.105 |
| >3 | 7.31 | 3.49–17.9 | <0.001 | 1.96 | 1.29–3.02 | <0.001 | 0.67 | 0.36–1.23 | 0.197 | 0.89 | 0.59–1.35 | 0.578 | 1.40 | 0.86–2.29 | 0.177 |
| Emergency Case Designation | 1.04 | 0.75–1.42 | 0.817 | 0.88 | 0.67–1.15 | 0.361 | 1.18 | 0.80–1.70 | 0.395 | 1.01 | 0.77–1.32 | 0.922 | 0.98 | 0.72–1.33 | 0.922 |
3.2 Readmission
Compared to patients who did not receive preoperative transfusion, those who received preoperative transfusion had 1.98 (95 % CI 1.39 to 2.76, p < 0.001) times higher crude odds of readmission within 30 days post operation. After adjusting for confounding variables, multivariable regression analysis demonstrated 1.49 (95 % CI 1.03 to 2.13, p = 0.031) higher odds of readmission in patients who received a preoperative transfusion [Table 4].
3.3 Reoperation
Compared to patients who did not receive preoperative transfusion, those who received preoperative transfusion had 1.53 (95 % CI 0.84 to 2.58, p = 0.133) times higher unadjusted (crude) odds of reoperation within 30 days post operation. After adjusting for confounding variables, multivariable regression analysis did not demonstrate a higher odds of reoperation in patients who received a preoperative transfusion (95 % CI 0.77 to 2.60, p = 0.207) [Table 4].
3.4 Hemodynamic complications
Compared to patients who did not receive preoperative transfusion, those who received preoperative transfusion had 1.23 (95 % CI 0.74 to 1.93, p = 0.393) times higher crude odds of hemodynamic complications. After adjusting for confounding variables, multivariable regression analysis did not demonstrate a higher odds of hemodynamic complications in patients who received a preoperative transfusion (95 % CI 0.49 to 1.41, p = 0.571) [Table 4].
3.5 Mortality
Compared to patients who did not receive preoperative transfusion, those who received preoperative transfusion had 3.25 (95 % CI 2.28 to 4.55, p < 0.001) times higher crude odds of mortality within 30 days post operation. After adjusting for confounding variables, multivariable regression analysis demonstrated 1.73 (95 % CI 1.15 to 2.55, p = 0.007) higher odds of mortality in patients who received a preoperative transfusion [Table 4].
4 Discussion
Long bone fractures have been reported to result in substantial blood loss. This is particularly true for FSF which can result in up to 3 L of blood loss, frequently requiring blood transfusions.8 While transfusions can be a vital lifesaving measure, they carry their own risks of complications. Therefore, understanding how blood transfusions impact the prognosis of patients sustaining FSF is essential for developing successful institutional protocols. Improved protocols would aid in preventing inappropriate transfusions and, in cases where blood transfusion is necessary, evaluation of pertinent risk factors would aid in risk stratification and follow-up measures. Therefore, the present study retrospectively evaluated a large sample of patients who underwent surgical fixation for FSF. The primary goal was to determine how receipt of preoperative RBC transfusion impacted the prognosis of the surgery by evaluating postoperative complications, including infection, hospital readmission, reoperation, hemodynamic complications, and mortality. The secondary goal was to determine factors that increased the risk of receiving a blood transfusion preoperatively. A better understanding of both of these goals would result in fewer patients receiving unnecessary transfusions, and better risk management for patients requiring transfusion.
Both cohorts in this study had a similar demographic, baseline health status, and procedure characteristic breakdown. The highest proportion of patients in this study were of older age and female, which is representative of the general population that suffers FSF – that being low energy trauma in patients with poor bone quality. Additionally, most patients did not have chronic conditions aside from hypertension, which is also representative of the general population over 40 years old.25
Our analysis revealed that increased age, history of bleeding disorders, partially or totally dependent preoperative functional status, ASA classification ≥3, and low hematocrit and WBC levels were associated with increased risk of a receiving preoperative transfusion, which is consistent with the general expectation considering that these factors represent a more volatile medical situation. It is worth noting that only age 90 years and older showed significantly increased risk for preoperative transfusion. Other demographic factors including black race and Hispanic ethnicity were also associated with increased risk for preoperative transfusion.
Prior studies measuring the effects of transfusion administration during orthopedic surgery are focused on total hip and knee arthroplasty, which are generally done in an elective setting. Additionally, much of the literature focused on intra- and post-operative transfusion as opposed to preoperative transfusion. Thus, this study is the first to our knowledge that focuses on preoperative transfusions in FSF, which is generally done in an urgent/emergent setting. The existing literature shows varying associations between transfusion and postoperative mortality, readmission, and reoperation with some studies in agreement with our results that transfusion is associated with increased risk of mortality and readmission,26–28 while other studies disagree.29,30 Owens et al. found an increased rate of mortality from 0.08 % in un-transfused patients to 0.56 % in transfused patients undergoing total knee arthroplasty (TKA) (p < 0.0001), while Pedersen et al. reported an odds ratio of 2.2 (95 % CI: 1.2–3.8) for 90-day mortality following total hip arthroplasty (THA) in transfused patients. Interestingly, Jans et al. aimed to characterize the mechanisms in which transfused patients died following total hip arthroplasty and described cases where mortality was directly attributable to transfusion-related acute lung injury (TRALI), transfusion-associated cardiac overload, and infection occurring after RBC transfusion.31 The differences of outcomes in the literature may reflect discrepancies in study design and incorporated variables and should be further explored.
Our analysis showed that patients who received preoperative transfusion had a significantly increased risk of infection. However, when studying rates of specific infection types, only UTI occurrence was shown to be significantly increased in the transfusion cohort, while pneumonia, sepsis, septic shock, superficial SSI, deep SSI, and PJI showed insignificant correlation to transfusion. There is support in the literature that transfusion can negatively impact immune function and increase infection rates in recipients, specifically superficial and deep SSI, sepsis, and pulmonary infections.15,17,18,32,33 One study found infection to be a significant cause of death in transfused patients postoperatively.31 In this case, the small sample size for each subgroup of infection in the transfusion cohort may be a limitation in this aspect of the analysis and more focus should be placed on predisposing factors of infection when designing institutional transfusion protocols. Alternatively, there may be confounding variables that increase the risk of UTI, such as a longer hospital length of stay, in the transfusion cohort as opposed to the transfusion itself being a predisposing factor.
Hemodynamic states were expected to differ between the patients receiving the transfusion versus those who did not receive the transfusion, considering that transfusion administration would seemingly be more likely in hemodynamically unstable patients. However, our analysis found no statistically significant differences in any hemodynamic complication rates between the two cohorts. Other studies27,29,30 found no differences in rates of thromboembolic events, myocardial infarction and stroke following perioperative transfusion during major orthopedic procedures.
The results of this study suggest that a more conservative strategy for transfusing patients with FSF who are otherwise stable may yield superior postoperative outcomes. The current consensus for preoperative transfusion in patients undergoing orthopedic surgeries — supported by the Association for the Advancement of Blood & Biotherapies (AABB) 2023 international guidelines, as well as the American College of Cardiology 2024 perioperative management guideline for noncardiac surgery — is to consider transfusion when the hemoglobin level drops below 8 g/dL.34,35 “However, a prospective randomized study by Mullis et al. found that while functional outcomes were unchanged between orthopedic trauma patients transfused at thresholds of 5.5 g/dL and 7.0 g/dL, the 5.5 g/dL group experienced a significantly lower infection rate.36 Therefore, further research into the efficacy of various transfusion thresholds on surgical outcomes and complications of FSF is warranted.
This study had several strengths. The chosen study design was favored because it ensured that potentially confounding interactions could be controlled for between preoperative risk factors for specific postoperative complications, independent of preoperative transfusion. Additionally, the inclusion of multiple years of NSQIP data allowed for an increased sample size and the ability to control for varying transfusion protocols across institutions over time. There are, however, limitations to this study. Most of the limitations present in this study are common in large database research. There may be patient factors that were not designated as variables on the database and therefore could not be factored. Additionally, accuracy of the data and analysis is dependent on accurate coding.
While almost 35 % of patients undergoing surgical fixation for FSF were transfused with at least one unit of packed RBCs perioperatively, only 15.5 % of those patients received their transfusion preoperatively (5.48 % of total patients undergoing surgical fixation for FSF). Further studies should assess the rates of similar complications to those assessed here, in FSF patients receiving intra- and post-operative transfusions and determine whether there is a significant difference in predisposing factors or outcomes between FSF patients who received a transfusion preoperatively versus intra- and post-operatively.
5 Conclusion
While preoperative blood transfusion can be a vital lifesaving measure in the management of patients with femoral shaft fractures, they carry their own risks which should be thoroughly evaluated when designing institutional protocols for blood care during surgery. Our analysis showed an increased risk of postoperative mortality, readmission, and urinary tract infection, but no statistically significant increased risk of hemodynamic complications and reoperation in patients receiving preoperative transfusion for FSF. These findings can help guide surgeons as they balance optimal hemoglobin levels perioperatively and risks for postoperative complications that may be associated with the blood transfusions.
Guardian/patient consent
Patient consent was not required as all data were de-identified, aggregated, and publicly available.
Ethical statement
Ethical approval was not required for this study as it involved analysis of publicly available, de-identified data.
Credit author statement
Conceptualization: D.H.M
Methodology: D.H.MB.B.ZJ.M.D
Data curation: D.H.M., B.B.Z.
Formal analysis: J.M.D.
Writing – original draft: J.M.D.
Writing – review & editing: P.K.T, D.H.M., B.B.Z, C.B.P, Q.N.
Supervision: A.M.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
References
- National data of 6409 Swedish inpatients with femoral shaft fractures: stable incidence between 1998 and 2004. Injury. 2009;40(3):304-308.
- [Google Scholar]
- Population-based epidemiology of femur shaft fractures. J Trauma Acute Care Surg. 2013;74(6):1516-1520.
- [Google Scholar]
- Estimating the global incidence of femoral fracture from road traffic collisions: a literature review. J Bone Joint Surg Am. 2015;97(6)
- [Google Scholar]
- Femoral shaft fractures in eldery patients - an epidemiological risk analysis of incidence, mortality and complications. Injury. 2023;54(7)
- [Google Scholar]
- Femoral Shaft Fractures. Statpearls. Treasure Island (FL): Statpearls Publishing Copyright © 2024. 2024
- [Google Scholar]
- Epidemiology of fractures in 15,000 adults: the influence of age and gender. J Bone Joint Surg Br. 1998;80(2):243-248.
- [Google Scholar]
- Atypical femoral fractures and bisphosphonate use: current evidence and clinical implications. Ther Adv Chronic Dis. 2015;6(4):185-193.
- [Google Scholar]
- Investigation of perioperative blood loss of femoral shaft fractures treated with intramedullary nail or locking compression plate. Injury. 2021;52(7):1891-1896.
- [Google Scholar]
- Clinical analysis of peri-operative hidden blood loss of elderly patients with intertrochanteric fractures treated by unreamed proximal femoral nail anti-rotation. Sci Rep. 2018;8(1):3225.
- [Google Scholar]
- Anaemia impedes functional mobility after hip fracture surgery. Age Ageing. 2008;37(2):173-178.
- [Google Scholar]
- The effect of perioperative anemia on clinical and functional outcomes in patients with hip fracture. J Orthop Trauma. 2004;18(6):369-374.
- [Google Scholar]
- Anemia and patient blood management in hip and knee surgery: a systematic review of the literature. Anesthesiology. 2010;113(2):482-495.
- [Google Scholar]
- Transfusion in orthopaedic surgery : a retrospective multicentre cohort study. Bone Jt Open. 2021;2(10):850-857.
- [Google Scholar]
- Transfusion reactions: prevention, diagnosis, and treatment. Lancet. 2016;388(10061):2825-2836.
- [Google Scholar]
- Deleterious clinical effects of transfusion-associated immunomodulation: fact or fiction? Blood. 2001;97(5):1180-1195.
- [Google Scholar]
- The ONTraC Ontario program in blood conservation. Transfus Apher Sci. 2014;50(1):32-36.
- [Google Scholar]
- Allogeneic blood transfusion increases the risk of postoperative bacterial infection: a meta-analysis. J Trauma. 2003;54(5):908-914.
- [Google Scholar]
- Blood transfusion and risk of infection: new convincing evidence. JAMA. 2014;311(13):1293-1294.
- [Google Scholar]
- Transfusion trigger trial for functional outcomes in cardiovascular patients undergoing surgical hip fracture repair (FOCUS) Transfusion. 2006;46(12):2192-2206.
- [Google Scholar]
- Red blood cell transfusion for people undergoing hip fracture surgery. Cochrane Database Syst Rev. 2015;2015(4)
- [Google Scholar]
- Haemoglobin Concentrations for the Diagnosis of Anaemia and Assessment of Severity. 2011
- [Google Scholar]
- Guideline on Haemoglobin Cutoffs to Define Anaemia in Individuals and Populations. 2023
- [Google Scholar]
- Hypertension Prevalence Among Adults Aged 18 and Over: United States, 2017–2018. 2020
- [Google Scholar]
- Risk factors for post-operative blood transfusion following total knee arthroplasty. Iowa Orthop J. 2020;40(1):69-73.
- [Google Scholar]
- Allogeneic blood transfusion and prognosis following total hip replacement: a population-based follow up study. BMC Musculoskelet Disord. 2009;10:167.
- [Google Scholar]
- Transfusion rates are increasing following total hip arthroplasty: risk factors and outcomes. J Arthroplast. 2013;28(8 Suppl):34-37.
- [Google Scholar]
- Blood transfusion in primary total hip and knee arthroplasty. Incidence, risk factors, and thirty-day complication rates. J Bone Joint Surg Am. 2014;96(23):1945-1951.
- [Google Scholar]
- Mortality, morbidity and related outcomes following perioperative blood transfusion in patients with major orthopaedic surgery: a systematic review. Transfus Med Hemother. 2018;45(5):355-367.
- [Google Scholar]
- Transfusion-related mortality after primary hip arthroplasty--an analysis of mechanisms and confounders. Vox Sang. 2012;103(4):301-308.
- [Google Scholar]
- Impact of allogenic packed red blood cell transfusion on nosocomial infection rates in the critically ill patient. Crit Care Med. 2002;30(10):2249-2254.
- [Google Scholar]
- Transfusion guidelines in adult spine surgery: a systematic review and critical summary of currently available evidence. Spine J. 2022;22(2):238-248.
- [Google Scholar]
- Red blood cell transfusion: 2023 AABB international guidelines. JAMA. 2023;330(19):1892-1902.
- [Google Scholar]
- AHA/ACC/ACS/ASNC/HRS/SCA/SCCT/SCMR/SVM guideline for perioperative cardiovascular management for noncardiac surgery: a report of the American college of cardiology/american heart association joint committee on clinical practice Guidelines. Circulation. 2024;150(19):e351-e442.
- [Google Scholar]
- Orthopaedic trauma and anemia: conservative versus liberal transfusion strategy: a prospective randomized study. J Orthop Trauma. 2024;38(1):18-24.
- [Google Scholar]