Abstract
Introduction
Despite high rates of transfusion reported among hip fracture patients in the perioperative period, the relationship between perioperative transfusions and VTE has not been thoroughly explored. Therefore, we used a national database to evaluate how perioperative transfusions among patients undergoing surgical management of hip fractures impacted 1) deep vein thrombosis (DVT) and 2) pulmonary embolism (PE) risk.
Methods
The Targeted Hip Fracture Database of the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) was queried for patients undergoing surgical management of hip fractures from 2016 to 2019. A multivariate logistic regression was conducted using various patient-specific variables to identify risk factors for DVT and PE. A nearest-neighbor propensity score matched (PSM) comparison between patients receiving and not receiving perioperative blood transfusions (1:1) was additionally conducted.
Results
Prior to our PSM, preoperative transfusions were not associated with DVT incidence (OR: 1.48, 95% CI: 0.80–2.50; p = 0.2). However, intra-operative/post-operative transfusions (OR: 1.26, 95% CI: 1.02–1.56; p = 0.00.30) as well as the receipt of both transfusion types (OR: 1.81, 95% CI: 1.10–2.81; p = 0.012) were associated with an increased risk of DVT. The latter of these findings remained significant following PSM (OR: 1.73, 95% CI: 1.04–2.73; p = 0.025). No relationship was demonstrated between PE risk and perioperative transfusion receipt.
Conclusion
Our findings emphasize the importance of perioperative blood management strategies among patients undergoing surgical repair of hip fracture. Specifically, orthopaedic surgeons should aim to optimize hip fracture patients prior to surgical intervention as well as intra-operatively to reduce transfusion incidence.
Keywords: Transfusions, Hip fracture, Venous thromboembolism (VTE), Pulmonary embolism (PE), Deep vein thrombosis (DVT)
1. Introduction
Venous thromboembolism (VTE) is a common and serious complication experienced by patients suffering from hip fractures.1, 2, 3, 4 While the estimated prevalence of deep vein thrombosis (DVT) among this patient population has been reported between 11.1% and 32.8%,5, 6, 7 these rates can reach up to 62% if there is a delay of over 48 h between trauma and surgery.8 Various factors contribute to the incidence of this highly morbid and economically burdensome complication,9, 10, 11, 12 including the endothelial damage and hypercoagulation provoked by the traumatic injury.3 Subsequently, orthopaedic surgeons continue to evaluate perioperative risk factors and preventative strategies to reduce the rates of DVT and pulmonary embolism (PE) among hip fracture patients.2,13,14
In the current literature, commonly cited risk factors for VTE in hip fracture patients include advanced age and increased length of immobilization and hospitalization.6,13, 14, 15, 16 However, there remains limited evidence regarding the relationship between perioperative blood transfusions and VTE risk among this population.17 Specifically, a majority of currently available analyses report on small sample sizes or evaluate patients undergoing hip fracture management at a single institution.15,16,18 As recent studies have suggested that transfusion receipt may increase VTE risk among patients undergoing other lower extremity procedures, such as total joint arthroplasty (TJA), higher powered, multi-institutional studies may more adequately evaluate how VTE risk is impacted by transfusions received in the perioperative period.
Given the limitations in the current literature as well as the effort to reduce post-operative VTE incidence patients undergoing surgical correction of hip fractures, more information is needed regarding the relationship between perioperative blood transfusions and VTE risk. Therefore, the aim of this paper was to identify the association between perioperative blood transfusion and increased risk for (1) DVT and (2) PE after hip fracture surgery in a large, multi-institutional population.
2. Methods
2.1. Database
Data for all patients undergoing surgical management of hip fractures between 2016 and 2019 were collected from the Targeted Hip Fracture Database of the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP). The database contains over 150 variables, ranging from pre-operative to post-operative variables, from over 650 hospitals.19 Information regarding specific surgical procedures were collected from the ACS-NSQIP database with their corresponding Current Procedural Terminology (CPT) codes. This study did not require Institutional Review Board (IRB) approval as the data collected from ACS-NSQIP contained is anonymous.
2.2. Patient selection
All patients undergoing hip fracture surgery between 2016 and 2019 were identified. Patients were included in the study if the database contained information regarding their age, race, sex, body mass index (BMI), fracture pattern, surgical procedure performed, anesthesia type, patient comorbidity burden, length of stay (LOS), American Society of Anesthesiology (ASA) score, incidence of deep vein thrombosis (DVT), incidence of pulmonary embolism (PE), preoperative transfusion status and perioperative transfusion status (intra-/post-operative). A modified Charlson Comorbidity Index (CCI) Score was calculated based on variables available in NSQIP from 2016 to 2019. For patients between ages 50–60, one point was assigned; for patients between ages 60 and 70, two points were assigned; for patients between 70 and 80, three points were assigned, and for patients older than 80, four points were assigned. One point was assigned for history of congestive heart failure, one point for history of chronic obstructive pulmonary disease or dyspnea, one point for any form of diabetes, two points for dialysis or renal failure, three points for ascites, and six points for disseminated cancer.20 All missing values were imputed through multiple imputation by chained equation to avoid in the estimation of parameters, variable availability-based selection and the subsequent listwise deletion-associated bias.
2.3. Outcome measures
The primary outcome variables include the incidence of DVT and PE. The ACS-NSQIP database provides information regarding 30-day DVT and PE rates as part of its standardized data collection process.
2.4. Statistical analysis
Both univariate and multivariable logistic regressions were conducted to identify the variables that were associated with significantly increased DVT and PE risk. All variables collected were included in this analysis to minimize confounding effects. Afterwards, nearest neighbor propensity score matching (PSM) between patients that did and did not receive blood transfusions of any type (1:1) was performed to further evaluate the impact of blood transfusions on post-operative VTE. Patients were matched based on age, race, sex, BMI, ASA class, anesthesia type, smoking, hip fracture type, and modified CCI (LOS), at a caliper of 0.1. Both univariate and multivariable logistic regression analyses was then repeated on the matched cohort. A p-value <0.05 was considered significant. Statistical analyses were performed using R version 4.1.1 (R Project for Statistical Computing, Vienna, Austria).
3. Results
3.1. Pre-PSM patient demographics
A total of 46,274 patients were included in our analysis (Table 1). Among this cohort, 11,273 patients (24.36%) received perioperative blood transfusions. This included 867 patients received preoperative transfusions, 10,244 patients receiving intra-operative/post-operative transfusions, and 1029 patients receiving both types of perioperative transfusions. Of the 11,273 patients who underwent perioperative transfusion, 156 (1.4%) experienced DVT and 93 (0.8%) experienced PE.
Table 1.
Characteristics of all patients undergoing surgical repair of hip fractures between 2016 and 2019.
| Characteristic | No Transfusion, N = 34,134a | Preoperative, N = 867a | Intra-operative/Post-operative, N = 10,244a | Pre and Intra-operative/Post-operative, N = 1,029a | p-valueb |
|---|---|---|---|---|---|
| Age Category | <0.001 | ||||
| 18-29 | 123/34,134 (0.4%) | 1/867 (0.1%) | 10/10,244 (<0.1%) | 2/1029 (0.2%) | |
| 30-39 | 217/34,134 (0.6%) | 2/867 (0.2%) | 22/10,244 (0.2%) | 3/1029 (0.3%) | |
| 40-49 | 382/34,134 (1.1%) | 9/867 (1.0%) | 63/10,244 (0.6%) | 12/1029 (1.2%) | |
| 50-59 | 1521/34,134 (4.5%) | 26/867 (3.0%) | 265/10,244 (2.6%) | 31/1029 (3.0%) | |
| 60-69 | 4202/34,134 (12%) | 93/867 (11%) | 925/10,244 (9.0%) | 88/1029 (8.6%) | |
| 70-79 | 7878/34,134 (23%) | 178/867 (21%) | 1874/10,244 (18%) | 171/1029 (17%) | |
| 80-89 | 12,703/34,134 (37%) | 324/867 (37%) | 4157/10,244 (41%) | 428/1029 (42%) | |
| 90+ | 7108/34,134 (21%) | 234/867 (27%) | 2928/10,244 (29%) | 294/1029 (29%) | |
| Percent Female | 22,851/34,134 (67%) | 589/867 (68%) | 7655/10,244 (75%) | 682/1029 (66%) | |
| BMI | 25.40 (5.86) | 24.18 (5.73) | 24.64 (5.88) | 24.33 (5.89) | <0.001 |
| Intra-Op/Post-Op Transfusions | 0/34,134 (0%) | 0/867 (0%) | 10,244/10,244 (100%) | 1029/1029 (100%) | <0.001 |
| Pre-Operative Transfusions | 0/34,134 (0%) | 0/867 (0%) | 10,244/10,244 (100%) | 1029/1029 (100%) | <0.001 |
| Deep Vein Thrombosis | 312/34,134 (0.9%) | 13/867 (1.5%) | 136/10,244 (1.3%) | 20/1029 (1.9%) | <0.001 |
| Pulmonary Embolism | 264/34,134 (0.8%) | 4/867 (0.5%) | 86/10,244 (0.8%) | 7/1029 (0.7%) | 0.6 |
| Medical DVT Prophylaxis Continued 28 Days Post-Op | 20,987/34,134 (61%) | 561/867 (65%) | 6565/10,244 (64%) | 652/1029 (63%) | <0.001 |
| Race | <0.001 | ||||
| American Indian or Alaska Native | 241/34,134 (0.7%) | 6/867 (0.7%) | 80/10,244 (0.8%) | 9/1029 (0.9%) | |
| Asian | 677/34,134 (2.0%) | 32/867 (3.7%) | 252/10,244 (2.5%) | 42/1029 (4.1%) | |
| Black or African American | 1429/34,134 (4.2%) | 49/867 (5.7%) | 504/10,244 (4.9%) | 70/1029 (6.8%) | |
| Native Hawaiian or Pacific Islander | 45/34,134 (0.1%) | 1/867 (0.1%) | 12/10,244 (0.1%) | 2/1029 (0.2%) | |
| White, Hispanic | 869/34,134 (2.5%) | 38/867 (4.4%) | 325/10,244 (3.2%) | 34/1029 (3.3%) | |
| White, Non-Hispanic | 30,873/34,134 (90%) | 741/867 (85%) | 9071/10,244 (89%) | 872/1029 (85%) | |
| Functional Status | <0.001 | ||||
| Independent | 26,976/34,134 (79%) | 612/867 (71%) | 7653/10,244 (75%) | 724/1029 (70%) | |
| Partially Dependent | 6116/34,134 (18%) | 219/867 (25%) | 2245/10,244 (22%) | 260/1029 (25%) | |
| Totally Dependent | 1042/34,134 (3.1%) | 36/867 (4.2%) | 346/10,244 (3.4%) | 45/1029 (4.4%) | |
| Smokers | 4380/34,134 (13%) | 102/867 (12%) | 930/10,244 (9.1%) | 90/1029 (8.7%) | <0.001 |
| Total Length of Hospital Stay | 6.59 (6.01) | 9.22 (8.09) | 7.58 (6.06) | 9.17 (7.41) | <0.001 |
| CPT Code | <0.001 | ||||
| 27236 | 14,430/34,134 (42%) | 195/867 (22%) | 2207/10,244 (22%) | 183/1029 (18%) | |
| 27244 | 3980/34,134 (12%) | 116/867 (13%) | 1133/10,244 (11%) | 130/1029 (13%) | |
| 27245 | 15,724/34,134 (46%) | 556/867 (64%) | 6904/10,244 (67%) | 716/1029 (70%) | |
| Place of Residence at 30 Days Post-Op | <0.001 | ||||
| Expired | 1809/34,134 (5.3%) | 97/867 (11%) | 728/10,244 (7.1%) | 115/1029 (11%) | |
| Facility which was home | 3423/34,134 (10%) | 86/867 (9.9%) | 1156/10,244 (11%) | 104/1029 (10%) | |
| Home | 16,786/34,134 (49%) | 274/867 (32%) | 3767/10,244 (37%) | 315/1029 (31%) | |
| Separate acute care | 361/34,134 (1.1%) | 25/867 (2.9%) | 122/10,244 (1.2%) | 17/1029 (1.7%) | |
| Skilled care | 10,036/34,134 (29%) | 309/867 (36%) | 3875/10,244 (38%) | 414/1029 (40%) | |
| Still in hospital | 1004/34,134 (2.9%) | 44/867 (5.1%) | 358/10,244 (3.5%) | 51/1029 (5.0%) | |
| Unskilled facility | 715/34,134 (2.1%) | 32/867 (3.7%) | 238/10,244 (2.3%) | 13/1029 (1.3%) | |
| Fracture Type | <0.001 | ||||
| Femoral neck fracture (subcapital, Garden types 1 and 2)-undisplaced | 3616/34,134 (11%) | 55/867 (6.3%) | 466/10,244 (4.5%) | 40/1029 (3.9%) | |
| Femoral neck fracture (subcapital, Garden types 3 and 4)-displaced | 11,517/34,134 (34%) | 142/867 (16%) | 1714/10,244 (17%) | 135/1029 (13%) | |
| Intertrochanteric | 16,700/34,134 (49%) | 587/867 (68%) | 6735/10,244 (66%) | 704/1029 (68%) | |
| Other/cannot be determined | 659/34,134 (1.9%) | 22/867 (2.5%) | 256/10,244 (2.5%) | 27/1029 (2.6%) | |
| Subtrochanteric | 1642/34,134 (4.8%) | 61/867 (7.0%) | 1073/10,244 (10%) | 123/1029 (12%) | |
| Modified Charlson Comorbidity Index | 3.78 (1.48) | 4.45 (1.85) | 4.10 (1.41) | 4.40 (1.70) | <0.001 |
Mean (SD); n/N (%).
Kruskal-Wallis rank sum test; Pearson's Chi-squared test.
3.2. Pre-PSM VTE risk factors
Prior to PSM and considered separately, BMI classification, comorbidity, hip fracture type, transfusion timing, and site of anesthesia were potential risk factors for DVT while smoking history and comorbidity were potential risk factors for PE (Table 2).
Table 2.
Risk factors for 30-day deep vein thrombosis and pulmonary embolism (Pre-PSM/Univariate).
| Characteristic | Deep Vein Thrombosis |
Pulmonary Embolism |
||||
|---|---|---|---|---|---|---|
| ORa | 95% CIa | p-value | ORa | 95% CIa | p-value | |
| SEX | ||||||
| Female | – | – | – | – | ||
| Male | 0.95 | 0.78, 1.15 | 0.57 | 1.00 | 0.80, 1.25 | >0.99 |
| Non-binary | 0.00 | – | 0.97 | 0.00 | – | 0.97 |
| RACE | ||||||
| White, Non-Hispanic | – | – | – | – | ||
| American Indian or Alaska Native | 0.57 | 0.09, 1.77 | 0.43 | 1.59 | 0.49, 3.75 | 0.36 |
| Asian | 0.86 | 0.41, 1.57 | 0.66 | 0.53 | 0.16, 1.24 | 0.20 |
| Black or African American | 1.12 | 0.72, 1.66 | 0.58 | 1.49 | 0.95, 2.23 | 0.065 |
| Native Hawaiian or Pacific Islander | 0.00 | – | 0.95 | 0.00 | – | 0.95 |
| White, Hispanic | 0.99 | 0.54, 1.64 | 0.96 | 1.79 | 1.06, 2.84 | 0.020 |
| BMI Classification | ||||||
| Normal weight | – | – | – | – | ||
| Underweight | 0.66 | 0.50, 0.87 | 0.003 | 1.01 | 0.75, 1.35 | 0.92 |
| Overweight | 1.12 | 0.90, 1.40 | 0.31 | 1.20 | 0.92, 1.56 | 0.18 |
| Obese Class I | 1.15 | 0.84, 1.55 | 0.36 | 1.33 | 0.93, 1.87 | 0.11 |
| Obese Class II | 1.25 | 0.75, 1.98 | 0.36 | 1.24 | 0.65, 2.16 | 0.47 |
| Obese Class III | 1.21 | 0.62, 2.12 | 0.54 | 1.47 | 0.69, 2.74 | 0.26 |
| SMOKE | ||||||
| No | – | – | – | – | ||
| Yes | 0.78 | 0.57, 1.05 | 0.11 | 0.67 | 0.45, 0.96 | 0.036 |
| Modified Charlson Comorbidity Index (CCI) | 1.07 | 1.01, 1.13 | 0.020 | 1.16 | 1.09, 1.23 | <0.001 |
| Pre-Op Platelet Count | 1.00 | 1.00, 1.00 | 0.59 | 1.00 | 1.00, 1.00 | 0.10 |
| Fracture Classification | ||||||
| Femoral neck fracture (Subcapital, Garden types 1 and 2) - Undisplaced | – | – | – | – | ||
| Femoral neck fracture (Subcapital, Garden types 3 and 4) - Displaced | 1.46 | 0.97, 2.27 | 0.083 | 0.87 | 0.61, 1.27 | 0.46 |
| Intertrochanteric | 1.80 | 1.22, 2.75 | 0.004 | 0.77 | 0.55, 1.10 | 0.13 |
| Other/cannot be determined | 2.18 | 1.08, 4.19 | 0.022 | 0.44 | 0.13, 1.10 | 0.12 |
| Subtrochanteric | 2.52 | 1.56, 4.14 | <0.001 | 1.15 | 0.71, 1.84 | 0.57 |
| All Transfusions | ||||||
| No Transfusion | – | – | – | – | ||
| Intra-operative/Post-operative | 1.46 | 1.19, 1.78 | <0.001 | 1.09 | 0.85, 1.38 | 0.51 |
| Pre and Intra-operative/Post-operative | 2.15 | 1.32, 3.30 | 0.001 | 0.88 | 0.37, 1.73 | 0.74 |
| Preoperative | 1.65 | 0.90, 2.77 | 0.079 | 0.59 | 0.18, 1.40 | 0.30 |
| CPT Code | ||||||
| 27244 | – | – | – | – | ||
| 27236 | 1.16 | 0.82, 1.67 | 0.42 | 1.45 | 0.99, 2.17 | 0.063 |
| 27245 | 1.73 | 1.25, 2.45 | 0.001 | 1.36 | 0.95, 2.03 | 0.11 |
| Anesthesia Type | ||||||
| General | – | – | – | – | ||
| Epidural | 0.00 | – | 0.98 | 1.67 | 0.09, 7.56 | 0.61 |
| MAC/IV Sedation | 1.00 | 0.72, 1.35 | >0.99 | 1.04 | 0.70, 1.48 | 0.86 |
| Regional | 0.00 | – | 0.96 | 0.48 | 0.03, 2.16 | 0.47 |
| Spinal | 0.59 | 0.45, 0.76 | <0.001 | 0.98 | 0.76, 1.26 | 0.90 |
| Length Of Stay >2 Days | ||||||
| No | – | – | – | – | ||
| Yes | 0.95 | 0.67, 1.39 | 0.77 | 1.06 | 0.70, 1.70 | 0.80 |
| Location at 30-Days Post-Op | ||||||
| Expired | – | – | – | – | ||
| Facility which was home | 0.75 | 0.45, 1.27 | 0.28 | 0.26 | 0.16, 0.42 | <0.001 |
| Home | 0.89 | 0.60, 1.38 | 0.58 | 0.28 | 0.20, 0.39 | <0.001 |
| Separate acute care | 1.21 | 0.45, 2.77 | 0.67 | 0.50 | 0.17, 1.14 | 0.14 |
| Skilled care | 1.38 | 0.93, 2.13 | 0.13 | 0.47 | 0.34, 0.66 | <0.001 |
| Still in hospital | 2.65 | 1.60, 4.45 | <0.001 | 1.02 | 0.63, 1.60 | 0.95 |
| Unskilled facility | 1.17 | 0.55, 2.31 | 0.67 | 0.37 | 0.15, 0.76 | 0.013 |
OR = Odds Ratio, CI = Confidence Interval.
After considering all potential risk factors simultaneously, preoperative transfusions were not associated with the incidence of DVT (OR: 1.48, 95% CI: 0.80 to 2.50; p = 0.2) (Table 3). However, intra-operative/post-operative transfusion receipt (OR: 1.26, 95% CI: 1.02 to 1.56; p = 0.00.30) as well as the receipt of both pre- and intra-operative/post-operative transfusions (OR: 1.81, 95% CI: 1.10 to 2.81; p = 0.012) were associated with an increased risk of DVT.
Table 3.
Risk factors for 30-day deep vein thrombosis and pulmonary embolism (Pre-PSM/Multivariate).
| Characteristic | Deep Vein Thrombosis |
Pulmonary Embolism |
||||
|---|---|---|---|---|---|---|
| ORa | 95% CIa | p-value | ORa | 95% CIa | p-value | |
| SEX | ||||||
| Female | – | – | – | – | ||
| Male | 0.93 | 0.76, 1.14 | 0.5 | 0.95 | 0.75, 1.19 | 0.7 |
| Non-binary | 0.00 | >0.9 | 0.00 | >0.9 | ||
| RACE | ||||||
| White, Non-Hispanic | – | – | – | – | ||
| American Indian or Alaska Native | 0.55 | 0.09, 1.72 | 0.4 | 1.71 | 0.52, 4.09 | 0.3 |
| Asian | 0.89 | 0.42, 1.63 | 0.7 | 0.56 | 0.17, 1.31 | 0.2 |
| Black or African American | 1.08 | 0.70, 1.60 | 0.7 | 1.48 | 0.93, 2.21 | 0.076 |
| Native Hawaiian or Pacific Islander | 0.00 | 0.00, 866 | >0.9 | 0.00 | 0.00, 1.24 | >0.9 |
| White, Hispanic | 0.92 | 0.50, 1.54 | 0.8 | 1.82 | 1.07, 2.89 | 0.017 |
| BMI Classification | ||||||
| Normal weight | – | – | – | – | ||
| Underweight | 0.67 | 0.50, 0.88 | 0.005 | 0.95 | 0.70, 1.27 | 0.7 |
| Overweight | 1.11 | 0.89, 1.38 | 0.4 | 1.21 | 0.93, 1.58 | 0.2 |
| Obese Class I | 1.11 | 0.81, 1.49 | 0.5 | 1.35 | 0.94, 1.90 | 0.094 |
| Obese Class II | 1.16 | 0.69, 1.84 | 0.5 | 1.24 | 0.65, 2.17 | 0.5 |
| Obese Class III | 1.07 | 0.55, 1.90 | 0.8 | 1.44 | 0.68, 2.70 | 0.3 |
| SMOKE | ||||||
| No | – | – | – | – | ||
| Yes | 0.84 | 0.61, 1.13 | 0.3 | 0.73 | 0.49, 1.05 | 0.10 |
| Modified Charlson Comorbidity Index (CCI) | 1.03 | 0.97, 1.09 | 0.3 | 1.11 | 1.04, 1.18 | 0.001 |
| Pre-Op Platelet Count | 1.00 | 1.00, 1.00 | 0.8 | 1.00 | 1.00, 1.00 | 0.049 |
| Fracture Classification | ||||||
| Femoral neck fracture (Subcapital, Garden types 1 and 2) - Undisplaced | – | – | – | – | ||
| Femoral neck fracture (Subcapital, Garden types 3 and 4) - Displaced | 1.46 | 0.97, 2.29 | 0.084 | 0.83 | 0.58, 1.22 | 0.3 |
| Intertrochanteric | 1.34 | 0.84, 2.21 | 0.2 | 0.72 | 0.46, 1.16 | 0.2 |
| Other/cannot be determined | 1.71 | 0.83, 3.34 | 0.13 | 0.39 | 0.11, 0.99 | 0.076 |
| Subtrochanteric | 1.74 | 1.01, 3.08 | 0.050 | 1.01 | 0.57, 1.78 | >0.9 |
| All Transfusions | ||||||
| No Transfusion | – | – | – | – | ||
| Intra-operative/Post-operative | 1.26 | 1.02, 1.56 | 0.030 | 0.99 | 0.76, 1.27 | >0.9 |
| Pre and Intra-operative/Post-operative | 1.81 | 1.10, 2.81 | 0.012 | 0.72 | 0.30, 1.42 | 0.4 |
| Preoperative | 1.48 | 0.80, 2.50 | 0.2 | 0.48 | 0.15, 1.15 | 0.2 |
| CPT Code | ||||||
| 27244 | – | – | – | – | ||
| 27236 | 1.12 | 0.72, 1.77 | 0.6 | 1.26 | 0.78, 2.09 | 0.4 |
| 27245 | 1.51 | 1.09, 2.15 | 0.018 | 1.30 | 0.89, 1.95 | 0.2 |
| Anesthesia Type | ||||||
| General | – | – | – | – | ||
| Epidural | 0.00 | 0.00, 17.5 | >0.9 | 1.74 | 0.10, 7.97 | 0.6 |
| MAC/IV Sedation | 1.09 | 0.78, 1.48 | 0.6 | 1.01 | 0.68, 1.45 | >0.9 |
| Regional | 0.00 | 0.00, 0.00 | >0.9 | 0.48 | 0.03, 2.15 | 0.5 |
| Spinal | 0.64 | 0.49, 0.83 | <0.001 | 1.04 | 0.79, 1.34 | 0.8 |
| Length Of Stay >2 Days | ||||||
| No | – | – | – | – | ||
| Yes | 0.98 | 0.68, 1.47 | >0.9 | 1.22 | 0.79, 2.00 | 0.4 |
| Location at 30-Days Post-Op | ||||||
| Expired | – | – | – | – | ||
| Facility which was home | 0.83 | 0.50, 1.40 | 0.5 | 0.27 | 0.16, 0.43 | <0.001 |
| Home | 0.97 | 0.65, 1.51 | 0.9 | 0.29 | 0.21, 0.41 | <0.001 |
| Separate acute care | 1.34 | 0.50, 3.08 | 0.5 | 0.52 | 0.18, 1.19 | 0.2 |
| Skilled care | 1.38 | 0.93, 2.14 | 0.13 | 0.47 | 0.34, 0.66 | <0.001 |
| Still in hospital | 2.96 | 1.77, 5.02 | <0.001 | 1.10 | 0.67, 1.75 | 0.7 |
| Unskilled facility | 1.22 | 0.58, 2.43 | 0.6 | 0.36 | 0.15, 0.75 | 0.012 |
OR = Odds Ratio, CI = Confidence Interval.
All three transfusion types were not associated with PE risk after taking all potential risk factors into account (Table 3). Specifically, pre-operative (OR: 0.48, 95% CI: 0.15 to 1.15; p = 0.2), intra-operative/post-operative (OR: 0.99, 95% CI: 0.76 to 1.27; p > 0.9), and both transfusion types combined (OR: 0.72, 95% CI: 0.30 to 1.42, p = 0.2) were not associated with a significant change in PE incidence.
3.3. Post-PSM patient demographics
Following propensity score matching, a total of 12,121 patients that did not receive transfusions were included in our analysis (Table 4). Among patients not receiving transfusions, 133 (1.1%) and 99 (0.8%) suffered a post-operative DVT and PE, respectively. There were no significant differences in the breakdown of age category, BMI, or anesthesia technique among cohorts.
Table 4.
Patient characteristics following propensity score matching.
| Characteristic | No Transfusion, N = 12,121a | Preoperative, N = 865a | Intra-operative/Post-operative, N = 10,228a | Pre and Intra-operative/Post-operative, N = 1,028a | p-valueb |
|---|---|---|---|---|---|
| Age Category | 0.13 | ||||
| 18–29 | 6/12,121 (<0.1%) | 1/865 (0.1%) | 10/10,228 (<0.1%) | 2/1028 (0.2%) | |
| 30–39 | 22/12,121 (0.2%) | 2/865 (0.2%) | 22/10,228 (0.2%) | 3/1028 (0.3%) | |
| 40–49 | 66/12,121 (0.5%) | 9/865 (1.0%) | 63/10,228 (0.6%) | 12/1028 (1.2%) | |
| 50–59 | 293/12,121 (2.4%) | 26/865 (3.0%) | 265/10,228 (2.6%) | 31/1028 (3.0%) | |
| 60–69 | 1058/12,121 (8.7%) | 93/865 (11%) | 925/10,228 (9.0%) | 88/1028 (8.6%) | |
| 70–79 | 2193/12,121 (18%) | 178/865 (21%) | 1872/10,228 (18%) | 171/1028 (17%) | |
| 80–89 | 4951/12,121 (41%) | 324/865 (37%) | 4152/10,228 (41%) | 427/1028 (42%) | |
| 90+ | 3532/12,121 (29%) | 232/865 (27%) | 2919/10,228 (29%) | 294/1028 (29%) | |
| BMI Classification | 0.3 | ||||
| Normal weight | 5150/12,121 (42%) | 353/865 (41%) | 4356/10,228 (43%) | 458/1028 (45%) | |
| Obese Class I | 1053/12,121 (8.7%) | 73/865 (8.4%) | 858/10,228 (8.4%) | 91/1028 (8.9%) | |
| Obese Class II | 296/12,121 (2.4%) | 23/865 (2.7%) | 257/10,228 (2.5%) | 24/1028 (2.3%) | |
| Obese Class III | 203/12,121 (1.7%) | 8/865 (0.9%) | 182/10,228 (1.8%) | 19/1028 (1.8%) | |
| Overweight | 2740/12,121 (23%) | 182/865 (21%) | 2352/10,228 (23%) | 212/1028 (21%) | |
| Underweight | 2679/12,121 (22%) | 226/865 (26%) | 2223/10,228 (22%) | 224/1028 (22%) | |
| Percent Female | 8983/12,121 (74%) | 587/865 (68%) | 7640/10,228 (75%) | 681/1028 (66%) | <0.001 |
| Intra-Op/Post-Op Transfusions | 0/12,121 (0%) | 0/865 (0%) | 10,228/10,228 (100%) | 1028/1028 (100%) | <0.001 |
| Pre-Operative Transfusions | 0/12,121 (0%) | 865/865 (100%) | 0/10,228 (0%) | 1028/1028 (100%) | <0.001 |
| Pulmonary Embolism | 99/12,121 (0.8%) | 4/865 (0.5%) | 86/10,228 (0.8%) | 7/1028 (0.7%) | 0.6 |
| Deep Vein Thrombosis | 133/12,121 (1.1%) | 12/865 (1.4%) | 136/10,228 (1.3%) | 20/1028 (1.9%) | 0.072 |
| Medical DVT Prophylaxis Continued 28 Days Post-Op | 7294/12,121 (60%) | 559/865 (65%) | 6556/10,228 (64%) | 652/1028 (63%) | <0.001 |
| Race | 0.012 | ||||
| American Indian or Alaska Native | 70/12,121 (0.6%) | 6/865 (0.7%) | 80/10,228 (0.8%) | 9/1028 (0.9%) | |
| Asian | 327/12,121 (2.7%) | 31/865 (3.6%) | 248/10,228 (2.4%) | 41/1028 (4.0%) | |
| Black or African American | 590/12,121 (4.9%) | 49/865 (5.7%) | 504/10,228 (4.9%) | 70/1028 (6.8%) | |
| Native Hawaiian or Pacific Islander | 14/12,121 (0.1%) | 1/865 (0.1%) | 12/10,228 (0.1%) | 2/1028 (0.2%) | |
| White, Hispanic | 382/12,121 (3.2%) | 37/865 (4.3%) | 324/10,228 (3.2%) | 34/1028 (3.3%) | |
| White, Non-Hispanic | 10,738/12,121 (89%) | 741/865 (86%) | 9060/10,228 (89%) | 872/1028 (85%) | |
| Smokers | 1079/12,121 (8.9%) | 102/865 (12%) | 929/10,228 (9.1%) | 90/1028 (8.8%) | 0.040 |
| Principal Anesthesia Technique | 0.2 | ||||
| Epidural | 13/12,121 (0.1%) | 1/865 (0.1%) | 12/10,228 (0.1%) | 5/1028 (0.5%) | |
| General | 8820/12,121 (73%) | 615/865 (71%) | 7498/10,228 (73%) | 766/1028 (75%) | |
| Local | 0/12,121 (0%) | 0/865 (0%) | 1/10,228 (<0.1%) | 0/1028 (0%) | |
| MAC/IV Sedation | 789/12,121 (6.5%) | 58/865 (6.7%) | 684/10,228 (6.7%) | 49/1028 (4.8%) | |
| Other | 2/12,121 (<0.1%) | 0/865 (0%) | 2/10,228 (<0.1%) | 0/1028 (0%) | |
| Regional | 73/12,121 (0.6%) | 6/865 (0.7%) | 50/10,228 (0.5%) | 7/1028 (0.7%) | |
| Spinal | 2424/12,121 (20%) | 185/865 (21%) | 1981/10,228 (19%) | 201/1028 (20%) | |
| Total Length of Hospital Stay | 6.94 (6.62) | 9.21 (8.09) | 7.57 (6.06) | 9.16 (7.41) | <0.001 |
| Fracture Type | <0.001 | ||||
| Femoral neck fracture (subcapital, Garden types 1 and 2)-undisplaced | 570/12,121 (4.7%) | 55/865 (6.4%) | 466/10,228 (4.6%) | 40/1028 (3.9%) | |
| Femoral neck fracture (subcapital, Garden types 3 and 4)-displaced | 1979/12,121 (16%) | 142/865 (16%) | 1713/10,228 (17%) | 135/1028 (13%) | |
| Intertrochanteric | 8146/12,121 (67%) | 587/865 (68%) | 6734/10,228 (66%) | 704/1028 (68%) | |
| Other/cannot be determined | 298/12,121 (2.5%) | 22/865 (2.5%) | 256/10,228 (2.5%) | 27/1028 (2.6%) | |
| Subtrochanteric | 1128/12,121 (9.3%) | 59/865 (6.8%) | 1059/10,228 (10%) | 122/1028 (12%) | |
| Place of Residence at 30 Days Post-Op | <0.001 | ||||
| Expired | 831/12,121 (6.9%) | 97/865 (11%) | 726/10,228 (7.1%) | 114/1028 (11%) | |
| Facility which was home | 1336/12,121 (11%) | 86/865 (9.9%) | 1156/10,228 (11%) | 104/1028 (10%) | |
| Home | 4909/12,121 (40%) | 274/865 (32%) | 3764/10,228 (37%) | 315/1028 (31%) | |
| Separate acute care | 134/12,121 (1.1%) | 24/865 (2.8%) | 121/10,228 (1.2%) | 17/1028 (1.7%) | |
| Skilled care | 4229/12,121 (35%) | 308/865 (36%) | 3865/10,228 (38%) | 414/1028 (40%) | |
| Still in hospital | 385/12,121 (3.2%) | 44/865 (5.1%) | 358/10,228 (3.5%) | 51/1028 (5.0%) | |
| Unskilled facility | 297/12,121 (2.5%) | 32/865 (3.7%) | 238/10,228 (2.3%) | 13/1028 (1.3%) | |
| Modified Charlson Comorbidity Index | 4.17 (1.51) | 4.45 (1.86) | 4.10 (1.41) | 4.39 (1.69) | <0.001 |
Mean (SD); n/N (%).
Kruskal-Wallis rank sum test; Pearson's Chi-squared test.
3.4. Post-PSM VTE risk factors
After PSM and considered separately, BMI classification, transfusion timing, and site of anesthesia were potential risk factors for DVT while BMI classification, comorbidity, and pre-op platelet count were potential risk factors for PE (Table 5).
Table 5.
Post-PSM (univariate) risk factors for deep vein thrombosis and pulmonary embolism.
| Characteristic | Deep Vein Thrombosis |
Pulmonary Embolism |
||||
|---|---|---|---|---|---|---|
| ORa | 95% CIa | p-value | ORa | 95% CIa | p-value | |
| SEX | ||||||
| Female | – | – | – | – | ||
| Male | 0.87 | 0.66, 1.13 | 0.30 | 0.96 | 0.69, 1.32 | 0.83 |
| RACE | ||||||
| White, Non-Hispanic | – | – | – | – | ||
| American Indian or Alaska Native | 0.97 | 0.16, 3.05 | 0.96 | 0.79 | 0.04, 3.53 | 0.81 |
| Asian | 0.49 | 0.15, 1.16 | 0.16 | 0.80 | 0.25, 1.90 | 0.66 |
| Black or African American | 1.26 | 0.76, 1.95 | 0.34 | 1.72 | 0.99, 2.79 | 0.039 |
| Native Hawaiian or Pacific Islander | 0.00 | – | 0.97 | 0.00 | – | 0.97 |
| White, Hispanic | 0.82 | 0.37, 1.55 | 0.58 | 1.68 | 0.83, 3.03 | 0.11 |
| BMI Classification | ||||||
| Normal weight | – | – | – | – | ||
| Underweight | 0.68 | 0.48, 0.93 | 0.019 | 0.88 | 0.59, 1.29 | 0.52 |
| Overweight | 1.07 | 0.81, 1.41 | 0.63 | 1.12 | 0.77, 1.59 | 0.55 |
| Obese Class I | 0.78 | 0.49, 1.21 | 0.29 | 1.04 | 0.60, 1.72 | 0.87 |
| Obese Class II | 0.99 | 0.44, 1.90 | 0.98 | 1.28 | 0.49, 2.70 | 0.57 |
| Obese Class III | 0.72 | 0.22, 1.71 | 0.52 | 2.18 | 0.91, 4.43 | 0.049 |
| SMOKE | ||||||
| No | – | – | – | – | ||
| Yes | 0.67 | 0.41, 1.04 | 0.10 | 1.14 | 0.69, 1.77 | 0.58 |
| Modified Charlson Comorbidity Index (CCI) | 1.04 | 0.97, 1.12 | 0.24 | 1.12 | 1.03, 1.21 | 0.005 |
| Pre-Op Platelet Count | 1.00 | 1.00, 1.00 | 0.85 | 1.00 | 1.00, 1.00 | 0.044 |
| Fracture Classification | ||||||
| Femoral neck fracture (Subcapital, Garden types 1 and 2) - Undisplaced | – | – | – | – | ||
| Femoral neck fracture (Subcapital, Garden types 3 and 4) - Displaced | 1.17 | 0.64, 2.30 | 0.64 | 0.97 | 0.50, 2.07 | 0.93 |
| Intertrochanteric | 1.14 | 0.67, 2.17 | 0.65 | 0.86 | 0.47, 1.75 | 0.65 |
| Other/cannot be determined | 1.41 | 0.57, 3.36 | 0.44 | 0.37 | 0.06, 1.42 | 0.20 |
| Subtrochanteric | 1.40 | 0.74, 2.82 | 0.32 | 1.29 | 0.64, 2.81 | 0.49 |
| All Transfusions | ||||||
| No Transfusion | – | – | – | – | ||
| Intra-operative/Post-operative | 1.21 | 0.95, 1.55 | 0.11 | 1.03 | 0.77, 1.38 | 0.84 |
| Pre and Intra-operative/Post-operative | 1.79 | 1.08, 2.81 | 0.016 | 0.83 | 0.35, 1.67 | 0.64 |
| Preoperative | 1.27 | 0.66, 2.20 | 0.43 | 0.56 | 0.17, 1.35 | 0.26 |
| CPT Code | ||||||
| 27244 | – | – | – | – | ||
| 27236 | 1.47 | 0.93, 2.38 | 0.11 | 1.47 | 0.85, 2.65 | 0.18 |
| 27245 | 1.71 | 1.15, 2.65 | 0.012 | 1.60 | 0.99, 2.75 | 0.069 |
| Anesthesia Type | ||||||
| General | – | – | – | – | ||
| Epidural | 0.00 | – | 0.99 | 3.82 | 0.21, 18.0 | 0.19 |
| MAC/IV Sedation | 1.19 | 0.77, 1.77 | 0.41 | 0.66 | 0.31, 1.22 | 0.22 |
| Regional | 0.00 | – | 0.97 | 0.00 | – | 0.97 |
| Spinal | 0.62 | 0.44, 0.86 | 0.006 | 0.80 | 0.54, 1.14 | 0.23 |
| Length Of Stay >2 Days | ||||||
| Yes | 0.80 | 0.50, 1.38 | 0.38 | 0.83 | 0.46, 1.69 | 0.58 |
| Location at 30-Days Post-Op | ||||||
| Expired | – | – | – | – | ||
| Facility which was home | 0.99 | 0.55, 1.83 | 0.97 | 0.24 | 0.12, 0.45 | <0.001 |
| Home | 0.98 | 0.60, 1.67 | 0.92 | 0.29 | 0.19, 0.46 | <0.001 |
| Separate acute care | 1.67 | 0.55, 4.22 | 0.31 | 0.18 | 0.01, 0.83 | 0.090 |
| Skilled care | 1.46 | 0.91, 2.47 | 0.14 | 0.46 | 0.31, 0.70 | <0.001 |
| Still in hospital | 2.87 | 1.55, 5.38 | <0.001 | 1.09 | 0.59, 1.94 | 0.78 |
| Unskilled facility | 0.85 | 0.28, 2.13 | 0.74 | 0.55 | 0.21, 1.23 | 0.18 |
OR = Odds Ratio, CI = Confidence Interval.
Considering all potential risk factors simultaneously, pre-operative (OR: 1.29, 95% CI: 0.67 to 2.25; p = 0.4) and intra-operative/post-operative (OR: 1.17, 95% CI: 0.92 to 1.49; p = 0.2) transfusion receipt individually were not associated with an increased risk of DVT following our PSM (Table 6). However, the receipt of both types of transfusion remained significantly associated with an increased risk of DVT (OR: 1.73, 95% CI: 1.04 to 2.73; p = 0.025). No transfusion categories were associated with PE incidence (all p-values >0.05) (Table 6).
Table 6.
Post-PSM (multivariate) risk factors for deep vein thrombosis and pulmonary embolism.
| Characteristic | Deep Vein Thrombosis |
Pulmonary Embolism |
||||
|---|---|---|---|---|---|---|
| ORa | 95% CIa | p-value | ORa | 95% CIa | p-value | |
| SEX | ||||||
| Female | – | – | – | – | ||
| Male | 0.83 | 0.63, 1.09 | 0.2 | 0.89 | 0.63, 1.23 | 0.5 |
| RACE | ||||||
| White, Non-Hispanic | – | – | – | – | ||
| American Indian or Alaska Native | 0.89 | 0.15, 2.86 | 0.9 | 0.84 | 0.05, 3.86 | 0.9 |
| Asian | 0.51 | 0.16, 1.20 | 0.2 | 0.92 | 0.28, 2.19 | 0.9 |
| Black or African American | 1.26 | 0.76, 1.97 | 0.3 | 1.68 | 0.96, 2.74 | 0.052 |
| Native Hawaiian or Pacific Islander | 0.00 | 0.00, 23,506,506,724 | >0.9 | 0.00 | 0.00, 4,808,588,896 | >0.9 |
| White, Hispanic | 0.81 | 0.37, 1.54 | 0.6 | 1.69 | 0.83, 3.06 | 0.11 |
| BMI Classification | ||||||
| Normal weight | – | – | – | – | ||
| Underweight | 0.68 | 0.48, 0.94 | 0.024 | 0.85 | 0.56, 1.25 | 0.4 |
| Overweight | 1.02 | 0.77, 1.35 | 0.9 | 1.12 | 0.77, 1.60 | 0.5 |
| Obese Class I | 0.73 | 0.45, 1.13 | 0.2 | 1.02 | 0.58, 1.68 | >0.9 |
| Obese Class II | 0.88 | 0.39, 1.69 | 0.7 | 1.21 | 0.47, 2.58 | 0.7 |
| Obese Class III | 0.63 | 0.19, 1.51 | 0.4 | 2.01 | 0.83, 4.12 | 0.084 |
| SMOKE | ||||||
| No | – | – | – | – | ||
| Yes | 0.68 | 0.41, 1.07 | 0.12 | 1.19 | 0.72, 1.86 | 0.5 |
| Modified Charlson Comorbidity Index (CCI) | 1.03 | 0.95, 1.10 | 0.5 | 1.08 | 0.99, 1.17 | 0.073 |
| Pre-Op Platelet Count | 1.00 | 1.00, 1.00 | 0.7 | 1.00 | 1.00, 1.00 | 0.039 |
| Fracture Classification | ||||||
| Femoral neck fracture (Subcapital, Garden types 1 and 2) - Undisplaced | – | – | – | – | ||
| Femoral neck fracture (Subcapital, Garden types 3 and 4) - Displaced | 1.24 | 0.68, 2.46 | 0.5 | 0.98 | 0.50, 2.11 | >0.9 |
| Intertrochanteric | 1.02 | 0.54, 2.12 | >0.9 | 0.73 | 0.35, 1.68 | 0.4 |
| Other/cannot be determined | 1.34 | 0.53, 3.25 | 0.5 | 0.31 | 0.05, 1.22 | 0.14 |
| Subtrochanteric | 1.24 | 0.60, 2.71 | 0.6 | 1.03 | 0.45, 2.51 | >0.9 |
| All Transfusions | ||||||
| No Transfusion | – | – | – | – | ||
| Intra-operative/Post-operative | 1.17 | 0.92, 1.49 | 0.2 | 0.99 | 0.74, 1.33 | >0.9 |
| Pre and Intra-operative/Post-operative | 1.73 | 1.04, 2.73 | 0.025 | 0.73 | 0.30, 1.46 | 0.4 |
| Preoperative | 1.29 | 0.67, 2.25 | 0.4 | 0.50 | 0.15, 1.20 | 0.2 |
| CPT Code | ||||||
| 27244 | – | – | – | – | ||
| 27236 | 1.26 | 0.70, 2.29 | 0.5 | 1.13 | 0.55, 2.39 | 0.7 |
| 27245 | 1.56 | 1.05, 2.44 | 0.037 | 1.50 | 0.92, 2.59 | 0.12 |
| Anesthesia Type | ||||||
| General | – | – | – | – | ||
| Epidural | 0.00 | 0.00, 1,375,817,556 | >0.9 | 3.76 | 0.21, 18.6 | 0.2 |
| MAC/IV Sedation | 1.20 | 0.77, 1.78 | 0.4 | 0.66 | 0.31, 1.23 | 0.2 |
| Regional | 0.00 | 0.00, 0.00 | >0.9 | 0.00 | 0.00, 0.00 | >0.9 |
| Spinal | 0.66 | 0.46, 0.91 | 0.016 | 0.85 | 0.57, 1.24 | 0.4 |
| Length Of Stay >2 Days | ||||||
| Yes | 0.92 | 0.55, 1.63 | 0.8 | 1.13 | 0.61, 2.34 | 0.7 |
| Location at 30-Days Post-Op | ||||||
| Expired | – | – | – | – | ||
| Facility which was home | 1.02 | 0.56, 1.90 | >0.9 | 0.25 | 0.12, 0.47 | <0.001 |
| Home | 1.02 | 0.62, 1.76 | >0.9 | 0.29 | 0.18, 0.45 | <0.001 |
| Separate acute care | 1.81 | 0.59, 4.60 | 0.2 | 0.18 | 0.01, 0.86 | 0.10 |
| Skilled care | 1.47 | 0.91, 2.51 | 0.13 | 0.45 | 0.30, 0.70 | <0.001 |
| Still in hospital | 3.14 | 1.67, 5.97 | <0.001 | 1.16 | 0.61, 2.10 | 0.6 |
| Unskilled facility | 0.87 | 0.29, 2.21 | 0.8 | 0.55 | 0.20, 1.23 | 0.2 |
OR = Odds Ratio, CI = Confidence Interval.
4. Discussion
With the significant adverse outcomes and financial burden associated with venous thromboembolism among hip fracture patients, the present analysis sought to better characterize the relationship between perioperative transfusions and the risk of post-operative PE and DVT. While controlling for various patient demographics, comorbidity burden, and fracture pattern, we found that intra-operative/post-operative transfusion receipt was associated with the incidence of 30-day DVT. Additionally, we found that the receipt of both pre-operative and intra-operative/post-operative transfusions was associated with a significantly higher risk of DVT. The latter of these findings remained significant following our PSM.
In their evaluation of 1007 patients suffering from proximal femur fractures, Johnston et al. found comparable rates of DVT/VTE among transfused (n = 28; 2.6%) and nontransfused (n = 55; 2.2%) patients (p = 0.52).18 This was similarly demonstrated by Feng et al. who failed to report any difference in transfusion rates between hip fracture patients suffering from DVT and those without post-operative VTE (p = 0.904).15 However, these analyses were underpowered and failed to control for various perioperative variables that may influence VTE risk. Conversely, Wu et al. demonstrated that perioperative blood transfusions (OR: 1.782, 95% CI: 1.031 to 2.896; p = 0.038) was associated with higher post-operative DVT incidence among 569 patients undergoing surgery for femoral and pelvic fractures.16 Our analysis builds from these findings by reporting on a substantially larger sample size of patients undergoing hip fracture surgery nationally.
Although the rate of transfusion receipt among patients undergoing elective orthopaedic surgery has continued to decrease over contemporary time frames,21 transfusion rates among hip fracture patients have been reported up to 45%.22, 23, 24 Our analysis reported a more moderate rate of 26.24% (12,140/46,274) for the receipt of any type of transfusion. However, there are multiple factors that contribute to the higher rate of transfusions among this patient population. Importantly, patients suffering from hip fracture are often elderly and medically complex, managing various comorbidities that have been independently associated with transfusion receipt.25, 26, 27, 28 Additionally, these patients frequently may be transferred between hospital systems without adequate management of hemorrhagic shock or hematoma formation precipitated by their bone trauma.29,30 Therefore, as our analysis demonstrated an association for the receipt of both types of transfusions following our PSM, our findings emphasize the need for optimized preoperative transfusion reduction strategies.
These approaches should focus on both reducing bleeding attributed to the fracture itself as well as minimizing the amount of intra-operative blood loss. In addition to implementing more restrictive transfusion protocols,31, 32, 33 preoperatively discontinuing or altering the anticoagulation therapy of hip fracture patients remains a foundational portion of the American College of Chest Physicians (ACCP) VTE prevention strategy.34 Additionally, a large amount of literature continues to emphasize the need to minimize the interval between presentation and surgery for this patient population to limit the rates of various complications.35,36 Notably, both Mattisson et al. and Yang et al. demonstrated significant reductions in transfusion requirements if surgery was performed within 24 h of fracture.37,38 Intra-operatively, adjuvant anti-fibrinolytic agents, such as tranexamic acid (TXA), continue to be frequently implemented to help reduce perioperative bleeding among patients undergoing lower extremity orthopaedic surgery.39, 40, 41 In their meta-analysis of randomized controlled trials (RCTs), Qi et al. reported significantly lower transfusion rates (p = 0.003), intra-operative blood loss (p = 0.02), and post-operative blood loss (p = 0.01) among hip fracture patients receiving TXA.40 While these efforts collectively have the potential to reduce DVT incidence based on the findings of our analysis, they may additional reduce other complications associated with transfusion receipt such as mortality, prolonged LOS, higher costs of care, and infection.17,21,42
Our analysis is limited that we are unable to control for the various perioperative transfusion reduction strategies implemented by healthcare centers treating included hip fracture patients. However, as we reported on over 40,000 patients undergoing surgical management of hip fractures, our analysis likely captured a wide variety of approaches, and therefore, we believe our findings are likely generalizable. Given the limitations of the ACS-NSQIP database, we were unable to identify the type or dosing of VTE prophylaxis included patients were managed with. However, the Targeted Hip Fracture Database does provide information regarding whether these medications were continued for the month following surgery and therefore, our findings were able to control for this factor. While the ACS-NSQIP database provides 30-day follow-up for included patients, it is possible that we were unable to capture all episodes of VTE that occurred within our patient population. However, the ACCP estimates that symptomatic VTE most commonly occurs within the first 2 post-operative weeks following major orthopaedic surgery.34 Additionally, the hypercoagulability following limb injury has been estimated to last up to a month,43 we likely were able to capture a large majority of DVTs and PEs suffered by our patient population. However, future study is needed to evaluate whether perioperative transfusions impact VTE risk beyond this study period.
5. Conclusion
The findings of our analysis emphasize the importance of perioperative blood management strategies among patients undergoing surgical repair of hip fracture. Specifically, orthopaedic surgeons should aim to optimize hip fracture patients prior to surgical intervention, as well as intra-operatively, to reduce transfusion incidence. Future analyses should evaluate the effect of various VTE prophylaxis regimens on hip fracture patients requiring transfusion in the perioperative period. Additionally, more information is needed regarding ideal transfusion cut-off values to reduce the rates of associated complications among this patient population.
Funding/sponsorship
None.
Institutional ethical committee approval
Institutional Review Board (IRB) was not required for our analysis as the NSQIP database contains de-identified data.
Authors contribution
Daniel Grits: Conceptualization, Methodology, Investigation, Formal Analysis, Writing Review & Editing, Andy Kuo: Methodology, Investigation, Formal Analysis, Writing Review & Editing, Alexander Acuña: Conceptualization, Investigation, Writing Original draft, Writing Review & Editing, Linsen T. Samuel: Conceptualization, Writing Review & Editing, Supervision, Atul F. Kamath: Conceptualization, Writing Review & Editing, Supervision, Administrative Support.
Declaration of competing interest
A.F.K. reports the following disclosures: research support (Signature Orthopaedics), paid presenter or speaker (DePuy Synthes and Zimmer Biomet), paid consultant (DePuy Synthes and Zimmer Biomet), stock or stock options (Zimmer Biomet, Johnson & Johnson, and Procter & Gamble), IP royalties (Innomed), and board or committee member (AAOS, AAHKS, and Anterior Hip Foundation). The other authors have nothing to disclose.
Acknowledgements
None.
References
- 1.Hefley W.F., Nelson C.L., Puskarich-May C.L. Effect of delayed admission to the hospital on the preoperative prevalence of deep-vein thrombosis associated with fractures about the hip. J Bone Joint Surg Am. 1996;78(4):581–583. doi: 10.2106/00004623-199604000-00012. [DOI] [PubMed] [Google Scholar]
- 2.Marsland D., Mears S.C., Kates S.L. Venous thromboembolic prophylaxis for hip fractures. Osteoporos Int. 2010;21(4) doi: 10.1007/s00198-010-1403-2. [DOI] [PubMed] [Google Scholar]
- 3.Rathbun S. The surgeon general's call to action to prevent deep vein thrombosis and pulmonary embolism. Circulation. 2009;119(15) doi: 10.1161/CIRCULATIONAHA.108.841403. [DOI] [PubMed] [Google Scholar]
- 4.Brill J.B., Badiee J., Zander A.L., et al. The rate of deep vein thrombosis doubles in trauma patients with hypercoagulable thromboelastography. J Trauma Acute Care Surg. 2017;83(3):413–419. doi: 10.1097/TA.0000000000001618. [DOI] [PubMed] [Google Scholar]
- 5.Tang Y., Wang K., Shi Z., Yang P., Dang X. A RCT study of Rivaroxaban, low-molecular-weight heparin, and sequential medication regimens for the prevention of venous thrombosis after internal fixation of hip fracture. Biomed Pharmacother. 2017;92:982–988. doi: 10.1016/j.biopha.2017.05.107. [DOI] [PubMed] [Google Scholar]
- 6.Song K., Yao Y., Rong Z., Shen Y., Zheng M., Jiang Q. The preoperative incidence of deep vein thrombosis (DVT) and its correlation with postoperative DVT in patients undergoing elective surgery for femoral neck fractures. Arch Orthop Trauma Surg. 2016;136(10):1459–1464. doi: 10.1007/s00402-016-2535-4. [DOI] [PubMed] [Google Scholar]
- 7.Shin W.C., Woo S.H., Lee S.J., Lee J.S., Kim C., Suh K.T. Preoperative prevalence of and risk factors for venous thromboembolism in patients with a hip fracture: an indirect multidetector CT venography study. J Bone Joint Surg Am. 2016;98(24):2089–2095. doi: 10.2106/JBJS.15.01329. [DOI] [PubMed] [Google Scholar]
- 8.Brauer C.A., Coca-Perraillon M., Cutler D.M., Rosen A.B. Incidence and mortality of hip fractures in the United States. JAMA. 2009;302(14):1573. doi: 10.1001/jama.2009.1462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ruppert A., Steinle T., Lees M. Economic burden of venous thromboembolism: a systematic review. J Med Econ. 2011;14(1):65–74. doi: 10.3111/13696998.2010.546465. [DOI] [PubMed] [Google Scholar]
- 10.Shahi A., Chen A.F., Tan T.L., Maltenfort M.G., Kucukdurmaz F., Parvizi J. The incidence and economic burden of in-hospital venous thromboembolism in the United States. J Arthroplasty. 2017;32(4):1063–1066. doi: 10.1016/j.arth.2016.10.020. [DOI] [PubMed] [Google Scholar]
- 11.Kim Y.-H., Oh S.H., Kim J.S. Incidence and natural history of deep-vein thrombosis after total hip arthroplasty. A prospective and randomised clinical study. J Bone Joint Surg Br. 2003;85(5):661–665. [PubMed] [Google Scholar]
- 12.Mahan C.E., Borrego M.E., Woersching A.L., et al. Venous thromboembolism: annualised United States models for total, hospital-acquired and preventable costs utilising long-term attack rates. Thromb Haemost. 2012;108(2):291–302. doi: 10.1160/TH12-03-0162. [DOI] [PubMed] [Google Scholar]
- 13.Xia Z.N., Xiao K., Zhu W., et al. Risk assessment and management of preoperative venous thromboembolism following femoral neck fracture. J Orthop Surg Res. 2018;13(1):1–9. doi: 10.1186/s13018-018-0998-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Zhao K., Zhang J., Li J., Meng H., Hou Z., Zhang Y. Incidence of and risk factors for new-onset deep venous thrombosis after intertrochanteric fracture surgery. Sci Rep. 2021;11(1):1–10. doi: 10.1038/s41598-021-96937-w. 2021 111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Feng L., Xu L., Yuan W., Xu Z., Feng Z., Zhang H. Preoperative anemia and total hospitalization time are the independent factors of preoperative deep venous thromboembolism in Chinese elderly undergoing hip surgery. BMC Anesthesiol. 2020;20(1):1–6. doi: 10.1186/s12871-020-00983-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wu L., Cheng B. Perioperative red blood cell infusion and deep vein thrombosis in patients with femoral and pelvic fractures: a propensity score matching. J Orthop Surg Res. 2021;16(1):1–10. doi: 10.1186/s13018-021-02510-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ponnusamy K.E., Kim T.J., Khanuja H.S. Perioperative blood transfusions in orthopaedic surgery. J Bone Jt Surg - Am. 2014;96(21):1836–1844. doi: 10.2106/JBJS.N.00128. [DOI] [PubMed] [Google Scholar]
- 18.Johnston P., Wynn-Jones H., Chakravarty D., Boyle A., Parker M.J. Is perioperative blood transfusion a risk factor for mortality or infection after hip fracture? J Orthop Trauma. 2006;20(10):675–679. doi: 10.1097/01.bot.0000249435.25751.e8. [DOI] [PubMed] [Google Scholar]
- 19.American College of Surgeons User guide for the 2017 ACS NSQIP participant use data file (PUF) Am Coll Surg. 2018;(October):1–24. [Google Scholar]
- 20.McConaghy K.M., Orr M.N., Grits D., Emara A.K., Molloy R.M., Piuzzi N.S. What is the 30-day mortality burden after elective total hip arthroplasty? An analysis of 194,062 patients. J Arthroplasty. 2021;36(10):3513–3518. doi: 10.1016/j.arth.2021.05.026. e2. [DOI] [PubMed] [Google Scholar]
- 21.Gu W.J., Gu X.P., Wu X.D., et al. Restrictive versus liberal strategy for red blood-cell transfusion: a systematic review and meta-analysis in orthopaedic patients. J Bone Joint Surg Am. 2018;100(8):686–695. doi: 10.2106/JBJS.17.00375. [DOI] [PubMed] [Google Scholar]
- 22.Cheung C.-L., Ang S Bin, Chadha M., et al. An updated hip fracture projection in Asia: the Asian Federation of Osteoporosis Societies study. Osteoporos sarcopenia. 2018;4(1):16–21. doi: 10.1016/j.afos.2018.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhang B.F., Wei X., Huang H., et al. Deep vein thrombosis in bilateral lower extremities after hip fracture: a retrospective study of 463 patients. Clin Interv Aging. 2018;13:681–689. doi: 10.2147/CIA.S161191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Arshi A., Lai W.C., Iglesias B.C., et al. Blood transfusion rates and predictors following geriatric hip fracture surgery. Hip Int. 2021;31(2):272–279. doi: 10.1177/1120700019897878. [DOI] [PubMed] [Google Scholar]
- 25.Pempe C., Werdehausen R., Pieroh P., et al. Predictors for blood loss and transfusion frequency to guide blood saving programs in primary knee- and hip-arthroplasty. Sci Rep. 2021;11(1):1–10. doi: 10.1038/s41598-021-82779-z. 2021 111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bian F.C., Cheng X.K., An Y.S. Preoperative risk factors for postoperative blood transfusion after hip fracture surgery: establishment of a nomogram. J Orthop Surg Res. 2021;16(1):1–8. doi: 10.1186/s13018-021-02557-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.White S.J.W., Cheung Z.B., Ye I., et al. Risk factors for perioperative blood transfusions in adult spinal deformity surgery. World Neurosurg. 2018;115:e731–e737. doi: 10.1016/j.wneu.2018.04.152. [DOI] [PubMed] [Google Scholar]
- 28.Tang J.H., Lyu Y., Cheng L.M., Li Y.C., Gou D.M. Risk factors for the postoperative transfusion of allogeneic blood in orthopedics patients with intraoperative blood salvage: a retrospective cohort study. Medicine (Baltim) 2016;95(8) doi: 10.1097/MD.0000000000002866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.O'Malley N.T., Blauth M., Suhm N., Kates S.L. Hip fracture management, before and beyond surgery and medication: a synthesis of the evidence. Arch Orthop Trauma Surg. 2011;131(11):1519–1527. doi: 10.1007/s00402-011-1341-2. [DOI] [PubMed] [Google Scholar]
- 30.Bateman L., Vuppala S., Porada P., et al. Medical management in the acute hip fracture patient: a comprehensive review for the internist. Ochsner J. 2012;12(2):101. [PMC free article] [PubMed] [Google Scholar]
- 31.Gregersen M., Borris L.C., Damsgaard E.M. Postoperative blood transfusion strategy in frail, anemic elderly patients with hip fracture: the TRIFE randomized controlled trial. Acta Orthop. 2015;86(3):363–372. doi: 10.3109/17453674.2015.1006980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Parker M.J. Randomised trial of blood transfusion versus a restrictive transfusion policy after hip fracture surgery. Injury. 2013;44(12):1916–1918. doi: 10.1016/j.injury.2013.04.033. [DOI] [PubMed] [Google Scholar]
- 33.American Academy of Orthopaedic Surgeons (AAOS). Hip Fractures in the Elderly - Clinical Practice Guideline | American Academy of Orthopaedic Surgeons.
- 34.Falck-Ytter Y., Francis C.W., Johanson N.A., et al. Prevention of VTE in orthopedic surgery patients. Antithrombotic therapy and prevention of thrombosis, 9th ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2012;141(2):e278S–e325S. doi: 10.1378/chest.11-2404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Pincus D., Ravi B., Wasserstein D., et al. Association between wait time and 30-day mortality in adults undergoing hip fracture surgery. JAMA. 2017;318(20):1994–2003. doi: 10.1001/jama.2017.17606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Borges F.K., Bhandari M., Guerra-Farfan E., et al. Accelerated surgery versus standard care in hip fracture (HIP ATTACK): an international, randomised, controlled trial. Lancet (London, England) 2020;395(10225):698–708. doi: 10.1016/S0140-6736(20)30058-1. [DOI] [PubMed] [Google Scholar]
- 37.Yang Y.-F., Huang J.-W., Gao X.-S., Liu Z.-L., Wang J.-W., Xu Z.-H. The correlation between timing of surgery and the need for RBC transfusions in the geriatric intertrochanteric fracture population. Geriatr Orthop Surg Rehabil. 2021;12 doi: 10.1177/2151459321998614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Mattisson L., Lapidus L.J., Enocson A. What is the influence of a delay to surgery >24 hours on the rate of red blood cell transfusion in elderly patients with intertrochanteric or subtrochanteric hip fractures treated with cephalomedullary nails? J Orthop Trauma. 2018;32(8):403–407. doi: 10.1097/BOT.0000000000001183. [DOI] [PubMed] [Google Scholar]
- 39.Haj-Younes B., Sivakumar B.S., Wang M., An V.V.G., Lorentzos P., Adie S. Tranexamic acid in hip fracture surgery: a systematic review and meta-analysis. J Orthop Surg. 2020;28(1) doi: 10.1177/2309499019887995. [DOI] [PubMed] [Google Scholar]
- 40.Qi Y ming, Wang H peng, Li Y juan, et al. The efficacy and safety of intravenous tranexamic acid in hip fracture surgery: a systematic review and meta-analysis. J Orthop Transl. 2019;19:1–11. doi: 10.1016/j.jot.2019.03.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Cheung Z.B., Anthony S.G., Forsh D.A., et al. Utilization, effectiveness, and safety of tranexamic acid use in hip fracture surgery: a population-based study. J Orthop. 2020;20:167–172. doi: 10.1016/j.jor.2020.01.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Maempel J.F., Wickramasinghe N.R., Clement N.D., Brenkel I.J., Walmsley P.J. The pre-operative levels of haemoglobin in the blood can be used to predict the risk of allogenic blood transfusion after total knee arthroplasty. Bone Jt J. 2016;98B(4):490–497. doi: 10.1302/0301-620X.98B4.36245. [DOI] [PubMed] [Google Scholar]
- 43.Meissner M.H., Chandler W.L., Elliott J.S. Venous thromboembolism in trauma: a local manifestation of systemic hypercoagulability? J Trauma. 2003;54(2):224–231. doi: 10.1097/01.TA.0000046253.33495.70. [DOI] [PubMed] [Google Scholar]