Traumaticsubdural hematoma (SDH) is one of the most common pathologies encountered in neurosurgical practice and is devastating for patients, with mortality rates reported as high as 90% in the literature.1,2Given the possibility of rapid clinical deterioration in patients with SDH, prompt surgical intervention is often warranted to mitigate further injury. Following surgical intervention, patients are often bedridden during the recovery process, raising concern for an increased risk of venous thromboembolism (VTE).3To prevent this event, VTE prophylactic agents such as heparin are often administered perioperatively.4,5In patients with SDH who undergo surgical intervention for clot evacuation and decompression, there is a concern that the resumption of prophylactic anticoagulation agents may increase the risk of secondary hemorrhage.6
The risk of postoperative VTE is exacerbated in patients with histories of anticoagulant use for hypercoagulable states. In these patients, anticoagulants are discontinued and exogenous vitamin K and coagulation factors are administered in emergency situations to minimize the risk of intra- and postoperative hemorrhage.7–9The clinical dilemma of whether to administer VTE prophylaxis (VTEP) is highlighted in these patients in the setting of SDH, as they are at an even greater risk of postoperative VTE. Despite this increased risk, the benefits versus risks of VTEP in this population have yet to be fully established in the literature.
The purpose of this study was to investigate the differences in hospital outcomes in patients with SDH and preexisting anticoagulant use who received VTEP following neurosurgical intervention compared with those who did not. Therefore, this study leverages data from the National Trauma Data Bank (NTDB) to conduct a propensity score–matched 3-year nationwide analysis of in-hospital mortality following SDH. The results of this study may better guide the development of postoperative anticoagulation guidelines in this patient population.
开云体育世界杯赔率
Patient Selection and Study Cohorts
In this retrospective cohort study, the NTDB was queried from 2017 to 2019 for all patients in the United States with a history of anticoagulant use presenting with a traumatic SDH subsequently requiring neurosurgical intervention (Committee on Trauma, American College of Surgeons. TQP PUF 2019.1.0, Chicago, IL, 2019). Patients with SDH were identified with ICD-10-CM codes matching the regular expression ^S06.5.*, and neurosurgical intervention including craniectomy was identified with ICD-10-PCS codes matching ^00.*|^0NB0.*. Patients who were missing data on VTEP or variables for propensity score matching, who received VTEP outside of the perioperative window, or who received VTEP too long after admission were excluded from the analyses. Cutoffs for the perioperative window and VTEP administration times were determined using the upper quartiles of each variable in hours, rounded to the nearest day. The remaining patients were allocated to the VTEP group, while patients who did not receive any prophylaxis in that time frame were allocated to the non-VTEP group (Fig. 1).
Study cohort selection using the NTDB 2017–2019 data.
Variables and Outcomes
Information about patient demographics, insurance type, injury severity, and preexisting comorbidities was collected. The evaluated demographics were patient age, sex, race, and ethnicity, and insurance types were Medicaid, Medicare, other government insurance, private insurance, or self-pay. The injury severity for each patient was determined using the Glasgow Coma Scale (GCS), highest Abbreviated Injury Scale (AIS) score, and presence of midline shift on brain imaging. AIS codes 140438–140446 and 140650–140656 were used to identify the severity of the SDH. Twenty-two different preexisting comorbidities were assessed, including hypertension, diabetes, cirrhosis, bleeding disorders, peripheral artery disease, history of cardiovascular accident, and history of myocardial infarction.
The primary outcome of interest was all-cause in-hospital mortality following surgery. Additionally, hospital length of stay (LOS) and complications such as pressure ulcers, catheter-associated urinary tract infection (UTI), stroke, pulmonary embolism, myocardial infarction, and unplanned reoperation were assessed. Discharge disposition such as discharges to home, home healthcare, rehabilitation, skilled nursing or intermediate care facility, and hospice were also assessed.
Propensity Score Matching
混杂因素对t的影响降到最低he effect of anticoagulant prophylaxis, a one-to-one propensity score–matching algorithm was used to identify non-VTEP patients with similar characteristics to the prophylaxis patients. Propensity scores were calculated for all patients based on demographics, insurance type, injury severity, comorbidities, and type of decompression surgery using a balanced logistic regression approach. One-to-one matching without replacement was performed using the k-nearest neighbors algorithm. All covariates were used for propensity score matching and are listed inTable 1.
Patient demographics, insurance type, injury severity, comorbidities, and neurosurgical treatment
| Variable | All, n = 2794 | Unmatched | Matched | |||
|---|---|---|---|---|---|---|
| Prophylaxis, n = 950 | Non-Prophylaxis, n = 1844 | p Value | Non-Prophylaxis, n = 950 | p Value | ||
| Demographics | ||||||
| Mean age ± SD, yrs | 74.05 ± 10.30 | 73.63 ± 10.43 | 74.26 ± 10.22 | 0.126 | 74.50 ± 9.89 | 0.069 |
| Female sex, n (%) | 953 (34.11) | 314 (33.05) | 639 (34.65) | 0.398 | 307 (32.3) | 0.732 |
| White, n (%) | 2357 (84.36) | 779 (82.00) | 1578 (85.57) | 0.014 | 821 (86.42) | <0.01 |
| Black, n (%) | 203 (7.27) | 92 (9.68) | 111 (6.02) | <0.001 | 65 (6.84) | 0.024 |
| Asian, n (%) | 90 (3.22) | 25 (2.63) | 65 (3.52) | 0.205 | 23 (2.42) | 0.770 |
| Native American, n (%) | 8 (0.29) | 4 (0.42) | 4 (0.22) | 0.339 | 2 (0.21) | 0.413 |
| Pacific Islander, n (%) | 10 (0.36) | 2 (0.21) | 8 (0.43) | 0.349 | 4 (0.42) | 0.413 |
| Hispanic/Latino, n (%) | 134 (4.80) | 40 (4.21) | 94 (5.10) | 0.299 | 36 (3.79) | 0.640 |
| Insurance type, n (%) | ||||||
| Medicaid | 100 (3.58) | 40 (4.21) | 60 (3.25) | 0.197 | 30 (3.16) | 0.223 |
| Medicare | 2052 (73.44) | 687 (72.32) | 1365 (74.02) | 0.333 | 714 (75.16) | 0.159 |
| Not billed | 2 (0.07) | 0 (0.00) | 2 (0.11) | 0.310 | 1 (0.11) | 0.317 |
| Government | 50 (1.79) | 14 (1.47) | 36 (1.95) | 0.366 | 15 (1.58) | 0.852 |
| Private | 478 (17.11) | 171 (18.00) | 307 (16.65) | 0.369 | 159 (16.74) | 0.467 |
| Self-pay | 31 (1.11) | 10 (1.05) | 21 (1.14) | 0.837 | 9 (0.95) | 0.818 |
| Injury severity | ||||||
| Mean GCS score ± SD | 11.26 ± 4.70 | 12.07 ± 4.27 | 10.84 ± 4.86 | <0.001 | 11.74 ± 4.51 | 0.112 |
| Mild 13–15, n (%) | 1738 (62.20) | 668 (70.32) | 1070 (58.03) | <0.001 | 639 (67.26) | 0.151 |
| Moderate 9–12, n (%) | 232 (8.30) | 72 (7.5) | 160 (8.68) | 0.319 | 67 (7.05) | 0.660 |
| Severe <8, n (%) | 824 (29.49) | 210 (22.11) | 614 (33.30) | <0.001 | 244 (25.68) | 0.067 |
| Mean AIS score ± SD | 4.61 ± 0.71 | 4.55 ± 0.75 | 4.64 ± 0.70 | <0.01 | 4.58 ± 0.74 | 0.270 |
| Midline shift, n (%) | 2057 (73.62) | 643 (67.68) | 1414 (76.68) | <0.001 | 676 (71.16) | 0.100 |
| Comorbidities, n (%) | ||||||
| Advanced directive | 308 (11.02) | 87 (9.16) | 221 (11.98) | 0.024 | 102 (10.74) | 0.250 |
| Alcoholism | 102 (3.65) | 25 (2.63) | 77 (4.18) | 0.039 | 30 (3.16) | 0.494 |
| 心绞痛 | 17 (0.61) | 6 (0.63) | 11 (0.60) | 0.910 | 5 (0.53) | 0.762 |
| Bleeding disorder | 107 (3.83) | 34 (3.58) | 73 (3.96) | 0.620 | 34 (3.58) | >0.99 |
| Cardiovascular accident | 363 (12.99) | 125 (13.16) | 238 (12.91) | 0.852 | 127 (13.37) | 0.892 |
| Chemotherapy | 32 (1.15) | 8 (0.84) | 24 (1.30) | 0.280 | 14 (1.47) | 0.198 |
| Cirrhosis | 40 (1.43) | 13 (1.37) | 27 (1.46) | 0.840 | 13 (1.37) | >0.99 |
| Congenital disorder | 18 (0.64) | 5 (0.53) | 13 (0.70) | 0.576 | 5 (0.53) | >0.99 |
| Congestive heart failure | 422 (15.10) | 149 (15.68) | 273 (14.80) | 0.539 | 148 (15.58) | 0.950 |
| COPD | 347 (12.42) | 130 (13.68) | 217 (11.77) | 0.146 | 116 (12.21) | 0.339 |
| Dementia | 200 (7.16) | 63 (6.63) | 137 (7.43) | 0.438 | 71 (7.47) | 0.473 |
| Diabetes | 957 (34.25) | 343 (36.11) | 614 (33.30) | 0.138 | 325 (34.21) | 0.387 |
| Disseminated cancer | 43 (1.54) | 8 (0.84) | 35 (1.90) | 0.032 | 8 (0.84) | >0.99 |
| Functional deficit | 488 (17.47) | 208 (21.89) | 280 (15.18) | <0.001 | 182 (19.16) | 0.140 |
| Hypertension | 2070 (74.09) | 724 (76.21) | 1346 (72.99) | 0.066 | 716 (75.37) | 0.668 |
| Myocardial infarction | 85 (3.04) | 35 (3.68) | 50 (2.71) | 0.156 | 26 (2.74) | 0.241 |
| Peripheral artery disease | 95 (3.40) | 38 (4.00) | 57 (3.09) | 0.209 | 31 (3.26) | 0.391 |
| Psychiatric disorder | 252 (9.02) | 84 (8.84) | 168 (9.11) | 0.814 | 84 (8.84) | >0.99 |
| Renal disease | 166 (5.94) | 54 (5.68) | 112 (6.07) | 0.680 | 57 (6.00) | 0.769 |
| Smoking | 259 (9.27) | 104 (10.95) | 155 (8.41) | 0.028 | 90 (9.47) | 0.289 |
| Steroid use | 58 (2.08) | 21 (2.21) | 37 (2.01) | 0.720 | 21 (2.21) | >0.99 |
| Substance use | 43 (1.54) | 15 (1.58) | 28 (1.52) | 0.902 | 13 (1.37) | 0.703 |
| Treatment | ||||||
| Mean op time ± SD, hrs | 86.07 ± 117.46 | 25.29 ± 44.06 | 117.38 ± 130.48 | <0.001 | 30.30 ± 44.17 | <0.01 |
| Craniotomy/craniectomy | 419 (15.00%) | 155 (16.32%) | 264 (14.32%) | 0.161 | 132 (13.89%) | 0.141 |
| Burr hole | 2053 (73.48%) | 680 (71.58%) | 1373 (74.46%) | 0.102 | 680 (71.58%) | >0.99 |
COPD = chronic obstructive pulmonary disease.
All variables were used for propensity-score matching. Boldface type indicates statistical significance; p values obtained by Pearson’s chi-square test or Student t-test.
Statistical Analysis
To assess for differences between the VTEP and non-VTEP groups, descriptive statistics were conducted using independent Student t-tests and Pearson’s chi-square tests before matching and paired Student t-tests and Pearson’s chi-square tests after propensity matching. The most common prophylactic agents given were identified and compared between living and dead patients.
Variables pertaining to patient demographics, injury severity, comorbidities, and VTEP agents used were evaluated as candidate covariates for multivariable regression analysis. The variance inflation factor (VIF) method was used to ensure all candidates had a VIF score < 5 to control for multicollinearity. Variables with VIF scores above the threshold were iteratively removed based on clinical significance. For each remaining variable, a univariable logistic regression using maximum likelihood estimation was developed with the selected variable as the independent variable and in-hospital death as the dependent variable. Using the variables that were significantly associated with death on univariable regression as the dependent variables, a multivariable model was then developed. Beta coefficients from the multivariable regression results were exponentiated to obtain odds ratios and confidence intervals.
The alpha level for significance was set at 0.05 for all analyses. Propensity matching and statistical analyses were performed in Python 3 using the PsmPy, SciPy, and statsmodels libraries.10–13
Results
Patient Selection and Study Cohorts
八千六百九十九名患者with SDH and a history of anticoagulant use undergoing one or more subsequent neurosurgical procedures were identified in the NTDB from 2017 to 2019. Of those patients, 5905 were excluded due to missing data, VTEP administered outside of the perioperative window, or VTEP administered more than 24 days after admission, resulting in a study cohort of 2794 patients. Nine hundred fifty of those patients received prophylaxis and 1844 did not. Following one-to-one propensity score matching, 950 matched non-VTEP controls were identified (Fig. 1).
Unmatched and Matched Group Differences
与无与伦比的non-VTEP组基地line, the VTEP group involved significantly fewer White (82% vs 85.57%, p = 0.014) and more Black (9.68% vs 6.02%, p < 0.001) patients, had higher mean GCS scores (12.07 vs 10.84, p < 0.001), lower mean AIS scores (4.55 vs 4.65, p < 0.01), lower rates of midline shift on imaging (67.68% vs 76.68%, p < 0.001), advanced directives (9.16% vs 11.98%, p = 0.024), alcohol use (2.63% vs 4.18%, p = 0.039), and disseminated cancer (0.84% vs 1.9%, p = 0.032), higher rates of functional deficit (21.89% vs 15.18%, p < 0.001) and smoking (10.95% vs 8.41%, p = 0.028), and mean earlier surgery (25.29 vs 117.38 hours from admission, p < 0.001). Following propensity score matching, most baseline differences were eliminated or significantly reduced (Table 1). The most common prophylactic agent used was unfractionated heparin (UH) received by 554 patients (58.32%), followed by low-molecular-weight heparin (LMWH) received by 365 patients (38.42%), other unspecified agent in 16 patients (1.68%), direct thrombin inhibitor in 11 patients (1.16%), and Xa inhibitor in 4 patients (0.42%). There were no significant differences in the agents used between living and dead patients (Table 2).
Time of VTEP administration and agent used
| Variable | All Pts, n = 950 | Alive, n = 774 | Dead, n = 176 | p Value* |
|---|---|---|---|---|
| Mean time of administration ± SD, hrs | ||||
| Time from admission | 75.22 ± 43.75 | 76.11 ± 44.92 | 71.28 ± 37.99 | 0.186 |
| Time from surgery | 49.92 ± 42.66 | 49.42 ± 42.52 | 52.13 ± 43.32 | 0.448 |
| Agent used, n (%) | ||||
| Direct thrombin inhibitor | 11 (1.16) | 7 (0.90) | 4 (2.27) | 0.126 |
| LMWH | 365 (38.42) | 306 (39.53) | 59 (33.52) | 0.139 |
| UH | 554 (58.32) | 445 (57.49) | 109 (61.93) | 0.281 |
| Xa inhibitor | 4 (0.42) | 2 (0.26) | 2 (1.14) | 0.104 |
| Other unspecified | 16 (1.68) | 14 (1.81) | 2 (1.14) | 0.531 |
pts = patients
p values obtained by Pearson’s chi-square or Student t-tests.
Of the evaluated outcomes after matching, the VTEP group had significantly lower rates of death (18.53% vs 34.53%, p < 0.001) but longer hospital LOS (14.09 vs 8.57 days, p < 0.001), as well as higher rates of pressure ulcers (2.11% vs 0.53%, p = 0.01), unplanned intensive care unit (ICU) admission (9.05% vs 3.47%, p < 0.001), and unplanned intubation (9.47% vs 6.11%, p = 0.021) compared with the matched non-VTEP group. VTEP group patients were also more likely to be discharged to rehabilitation (32.74% vs 19.68%, p < 0.001) and skilled nursing/intermediate care (25.58% vs 17.16%, p < 0.001;Table 3).
Outcomes, hospital complications, and discharge disposition
| Variable | All, n = 2794 | Unmatched | Matched | |||
|---|---|---|---|---|---|---|
| Prophylaxis, n = 950 | Non-Prophylaxis, n = 1844 | p Value* | Non-Prophylaxis, n = 950 | p Value* | ||
| Deaths, n (%) | 943 (33.75) | 176 (18.53) | 767 (41.59) | <0.001 | 328 (34.53) | <0.001 |
| Mean hospital LOS ± SD, days | 11.54 ± 9.53 | 14.09 ± 11.05 | 8.71 ± 6.40 | <0.001 | 8.57 ± 6.33 | <0.001 |
| Hospital complications, n (%) | ||||||
| Cardiac arrest | 65 (2.33) | 16 (1.68) | 49 (2.66) | 0.106 | 18 (1.89) | 0.795 |
| Catheter-associated UTI | 24 (0.86) | 9 (0.95) | 15 (0.81) | 0.716 | 5 (0.53) | 0.453 |
| Deep vein thrombosis | 61 (2.18) | 27 (2.84) | 34 (1.84) | 0.087 | 18 (1.89) | 0.322 |
| Kidney failure | 33 (1.18) | 16 (1.68) | 17 (0.92) | 0.077 | 11 (1.16) | 0.499 |
| Myocardial infarction | 29 (1.04) | 7 (0.74) | 22 (1.19) | 0.260 | 8 (0.84) | 0.795 |
| Organ space SSI | 3 (0.11) | 3 (0.32) | 0 (0.00) | 0.016 | 0 (0.00) | 0.181 |
| Pressure ulcer | 33 (1.18) | 20 (2.11) | 13 (0.70) | <0.01 | 5 (0.53) | 0.010 |
| Pulmonary embolism | 8 (0.29) | 5 (0.53) | 3 (0.16) | 0.088 | 3 (0.32) | 0.676 |
| Respiratory failure | 14 (0.50) | 5 (0.53) | 9 (0.49) | 0.892 | 2 (0.21) | 0.439 |
| Sepsis | 29 (1.04) | 17 (1.79) | 12 (0.65) | <0.01 | 6 (0.63) | 0.056 |
| Stroke/cardiovascular accident | 66 (2.36) | 23 (2.42) | 43 (2.33) | 0.883 | 19 (2.00) | 0.676 |
| Superficial incision SSI | 3 (0.11) | 2 (0.21) | 1 (0.05) | 0.232 | 1 (0.11) | 0.676 |
| Unplanned ICU admission | 139 (4.97) | 86 (9.05) | 53 (2.87) | <0.001 | 33 (3.47) | <0.001 |
| Unplanned intubation | 182 (6.51) | 90 (9.47) | 92 (4.99) | <0.001 | 58 (6.11) | 0.021 |
| Unplanned reoperation | 106 (3.79) | 41 (4.32) | 65 (3.52) | 0.300 | 36 (3.79) | 0.676 |
| Ventilator-associated pneumonia | 47 (1.68) | 21 (2.21) | 26 (1.41) | 0.119 | 11 (1.16) | 0.179 |
| Discharge disposition, n (%) | ||||||
| Home | 239 (8.55) | 86 (9.05) | 153 (8.30) | 0.499 | 106 (11.16) | 0.256 |
| Home healthcare | 159 (5.69) | 61 (6.42) | 98 (5.31) | 0.232 | 66 (6.95) | 0.738 |
| Rehabilitation | 576 (20.62) | 311 (32.74) | 265 (14.37) | <0.001 | 187 (19.68) | <0.001 |
| Skilled nursing/intermediate care | 603 (21.58) | 243 (25.58) | 360 (19.52) | <0.001 | 163 (17.16) | <0.001 |
| Hospice | 242 (8.66) | 65 (6.84) | 177 (9.60) | 0.014 | 93 (9.79) | 0.056 |
| Other | 28 (1.00) | 5 (0.53) | 23 (1.25) | 0.070 | 6 (0.63) | 0.795 |
SSI = surgical site infection.
Boldface type indicates statistical significance; p values obtained by Pearson’s chi-square or Student t-tests.
p values presented following Benjamini-Hochberg multiple comparisons correction.
Logistic Regression Analysis
使用VIF方法为共线控制ity, covariate candidates were identified with a maximum VIF of 1.25. Univariable logistic regression models were developed with 32 unique covariate candidates. A multivariable logistic regression model was then developed using age, sex, GCS score, AIS score, midline shift, medical and social comorbidities, and VTEP agents used as independent variables and in-hospital mortality as the dependent variable. The results of the multivariable logistic regression showed that UH (OR 0.36, p < 0.001) and LMWH (OR 0.3, p < 0.001) were associated with lower odds of in-hospital mortality. Additionally, higher GCS score (OR 0.83, p < 0.001) and dementia (OR 0.63, p = 0.017) were associated with lower odds of death, while older age (OR 1.02, p < 0.001), midline shift (OR 1.29, p = 0.026), peripheral artery disease (OR 2.07, p < 0.01), renal disease (OR 1.56, p = 0.019), advanced directives (OR 1.5, p < 0.01), and diabetes (OR 1.32, p < 0.01) were associated with higher odds of death (Fig. 2).
Logistic regression analysis for postoperative mortality prior to hospital discharge. Odds ratios are reported as the presence of a variable for binary variables or a higher value relative to the population for continuous variables. An OR > 1 represents greater odds of death and an OR < 1 represents lower odds of death. ADL = activity of daily living; COPD = chronic obstructive pulmonary disease; N/A = not applicable; * = significant.
Discussion
In this retrospective cohort study, the NTDB database was used to evaluate the effect of postoperative anticoagulant prophylaxis in patients with SDH and a history of anticoagulant use. One-to-one propensity score matching was used to control for the effect of potential confounders in demographics, injury characteristics, and hospital characteristics during descriptive statistics. Logistic regression analysis was used to evaluate the effect of various anticoagulation prophylaxis agents and other covariates on the risk of in-hospital mortality. The results showed that patients who received VTEP following neurosurgical intervention had lower mortality rates compared with patients who did not receive VTEP. Additionally, the VTEP group had a significantly longer LOS and increased rates of discharge to rehabilitation and skilled nursing/intermediate care when compared with the non-VTEP group, likely associated with the rates of complications including unplanned ICU admission, pressure ulcers, and intubation seen in the VTEP group.
Management of SDH
SDH can develop after traumatic brain injury due to mechanical forces disrupting the continuity of the brain and meningeal vasculature, resulting in cerebral edema and hematomas that lead to brain compression and ultimately raising the intracranial pressure.14–16The presence of an SDH is usually diagnosed through a CT scan of the head, indicated in patients with the appropriate risk factors or symptoms such as decreased GCS score, focal neurological deficits, and high-risk mechanisms of injury. Subsequently, urgent neurosurgical decompression is warranted to prevent high-risk sequelae such as brain herniation.17In patients receiving anticoagulants or antiplatelet agents, these medications are discontinued and anticoagulation reversal agents such as vitamin K, prothrombin complex concentrate, fresh frozen plasma, and platelets are administered prior to surgery to prevent a risk of hemorrhage expansion.18–20Postoperatively, clinicians face the dilemma of administering VTEP or bridging anticoagulation to prevent VTE versus withholding prophylaxis to reduce the risk of secondary hemorrhage. This dilemma is of particular concern in elderly patients, who are prone to cranial injury and hemorrhage due to falls and are heavily represented in this patient population.21
Mortality
All-cause mortality was used as the primary endpoint in our study because it reflects the worst-case sequelae of both intracranial hemorrhage secondary to VTEP as well as VTE-associated complications such as pulmonary embolism and embolic stroke. Accordingly, we found that the VTEP group had a significantly lower mortality rate compared with the non-VTEP group. Most patients receiving VTEP received UH or LMWH and these were the only agents of those assessed that were associated with a significantly reduced mortality rate. This finding may indicate that UH and LMWH are more effective in reducing mortality when compared with direct thrombin and factor Xa inhibitors, but this may also reflect the low sample sizes of patients who were treated with direct thrombin and factor Xa inhibitors. The effectiveness of heparin, especially LMWH, can be explained by its pharmacokinetics, with a 91% bioavailability and 275-minute half-life when injected subcutaneously.22Perioperative heparin was found to be safe to use during procedures such as tumor resection and subarachnoid hemorrhage evacuation, and had lower rates of intracranial hemorrhage compared with older anticoagulants such as warfarin in traumatic brain injury.23–26There is no consensus on the issue, however, as one randomized controlled study of 68 patients found that preoperative prophylactic enoxaparin increased the rate of postoperative intracranial hemorrhage.27,28While our study is not able to definitely conclude one decision is superior to the other, it does provide evidence toward the benefits of VTEP.
Postoperative Complications
In our study, the VTEP group had similar rates of reoperation compared with the non-VTEP group. Previous studies have shown that perioperative anticoagulant use was associated with an increased risk of hematoma expansion and intracranial hemorrhage in patients presenting with SDH.29,30This is the primary cause for concern when initiating antithrombotic therapy in patients with recent bleeding events, thus the lack of a difference in reoperation rates may indicate comparable rates of rebleeding.31In practice, the decision to initiate antithrombotic therapy is made when the risk of thromboembolic complications outweighs the risk of recurrent hemorrhage.32Our results also showed that the group receiving VTEP did not have higher rates of thromboembolic complications such as deep vein thrombosis, stroke, and pulmonary embolism either before or after matching. This finding may be explained by the inherently higher risk of thromboembolic complications in the VTEP group, indicating effective prophylaxis in preventing VTE.33
Increased LOS has been shown to be associated with increased rates of pressure ulcers, intubation, and ICU stay, and is commonly a concurrent finding in study cohorts with increased discharge to skilled nursing or intermediate care facilities, suggesting a common link between the significant complications found in the VTEP group.34–37The cause of these findings is unclear, but it is possible that patients are observed longer in the inpatient setting following VTEP due to the fear of subsequent hemorrhage, increasing the window of opportunity for additional complications. It is also possible that there was a higher rate of unspecified and unassociated complications in the VTEP group, leading to increased complications and subsequent longer LOS.
Additional Risk Factors of Mortality
Among the risk factors associated with mortality, the presence of midline shift, increased age, diabetes, presence of an advanced directive, renal disease, and peripheral artery disease were all associated with an increased risk of mortality. These findings are consistent with those in the current literature.38–41在这些患者的死亡风险更高,therapy with UH and LMWH may be especially beneficial despite the risk of increased hospital complications, due to the mortality benefit they confer. Conversely, higher GCS score and, surprisingly, a history of dementia were associated with lower odds of death. The former finding is relatively simple to understand, as better mental status is associated with lower injury severity and thus better survivability, but the latter finding is more confounding.42One potential explanation is that intracranial hemorrhage such as SDH is often on the differential diagnosis for vascular dementia, thus patients with a known history may receive more aggressive neuroimaging upon presentation.43The significance of this finding remains unclear and the connection between dementia and SDH is yet to be explored.
Limitations of the Study
The study must be viewed in the context of its limitations. The NTDB is a multisite database consisting of data from sites across the country, containing over 1 million data points per year. While this allows for large patient cohorts and high-powered analyses, it also increases the potential for errors in data collection, entry, and standardization, and often lacks hypothesis-specific data fields. For this reason, the study was unable to evaluate the risk of rebleeding as an outcome measure, which would have provided more insight into the lower rate of all-cause in-hospital mortality in the VTEP group. Furthermore, the NTDB is not a random sample of all hospitals across the country, so the patient cohort is inherently biased toward patient populations treated in the participant hospitals and may not necessarily be representative of the country. Lastly, other confounding factors may exist outside of those assessed in this study, such as coexisting injuries or long-bone fractures, that may play an explanatory role in the outcome variables assessed in this study.
Conclusions
患者创伤SDH和历史的ticoagulant use, VTEP was associated with reduced mortality but an increased rate of hospital complications and higher rates of discharge to rehabilitation and skilled nursing/intermediate care. UH and LMWH were the most common as well as the most effective agents used. In patients with a history of anticoagulant use suffering from traumatic SDHs, the prevention of VTE events may confer an in-hospital survival advantage despite the risk of secondary hemorrhage. Additional research is needed to further clarify this association and determine the optimal timing of VTEP to lower overall mortality rates following SDH in these patients.
Acknowledgments
The content reproduced from the Trauma Quality Program Participant Use File (TQP PUF) remains the full and exclusive copyrighted property of the American College of Surgeons. The American College of Surgeons is not responsible for any claims arising from works based on the original data, text, tables, or figures.
Disclosures
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Author Contributions
Conception and design: Mehta, Jiang, Chiu. Acquisition of data: Jiang. Analysis and interpretation of data: Mehta, Jiang, Hukamdad, Bhaskara, Chiu, Sadeh. Drafting the article: Mehta, Jiang, Hukamdad, Gould, Bhaskara, Sadeh. Critically revising the article: Mehta, Jiang, Hukamdad, Bhaskara, Chiu, Sadeh. Reviewed submitted version of manuscript: Mehta, Jiang, Gould, Bhaskara, Chiu, Sadeh. Approved the final version of the manuscript on behalf of all authors: Mehta. Statistical analysis: Jiang, Sadeh. Administrative/technical/material support: Mehta, Jiang, Hukamdad. Study supervision: Mehta, Sadeh.
References
-
1 ↑
AromatarioM,TorselloA,D’ErricoS,et al.Traumatic epidural and subdural hematoma: epidemiology, outcome, and dating.Medicina (Kaunas).2021;57(2):125.
-
2 ↑
HanifS,AbodundeO,AliZ,PidgeonC.Age related outcome in acute subdural haematoma following traumatic head injury.Ir Med J.2009;102(8):255–257.
-
3 ↑
SaketR,AggarwalS,KumarV,KumarP,PatelS.Acute venous thromboembolism in Indian patients of isolated proximal femur fractures.J Clin Orthop Trauma.2019;10(5):917–921.
-
4 ↑
EpsteinNE.A review of the risks and benefits of differing prophylaxis regimens for the treatment of deep venous thrombosis and pulmonary embolism in neurosurgery.Surg Neurol.2005;64(4):295–302.
-
5 ↑
GanauM,PriscoL,CebulaH,et al.Risk of Deep vein thrombosis in neurosurgery: state of the art on prophylaxis protocols and best clinical practices.J Clin Neurosci.2017;45:60–66.
-
6 ↑
García-RodríguezLA,GaistD,MortonJ,CooksonC,González-PérezA.Antithrombotic drugs and risk of hemorrhagic stroke in the general population.Neurology.2013;81(6):566–574.
-
7 ↑
HolleyAB,PetteysS,MitchellJD,HolleyPR,CollenJF.Thromboprophylaxis and VTE rates in soldiers wounded in Operation Enduring Freedom and Operation Iraqi Freedom.Chest.2013;144(3):966–973.
-
8
ShawJR,LiN,NixonJ,et al.Coagulation assays and direct oral anticoagulant levels among patients having an elective surgery or procedure.J Thromb Haemost.2022;20(12):2953–2963.
-
9 ↑
da SilvaIRF,FronteraJA.Resumption of anticoagulation after intracranial hemorrhage.Curr Treat Options Neurol.2017;19(11):39.
-
11
VirtanenP,GommersR,OliphantTE,et al.SciPy 1.0: fundamental algorithms for scientific computing in Python.Nat Methods.2020;17(3):261–272.
-
12
SeaboldS,PerktoldJ.Statsmodels: Econometric and statistical modeling with Python. In:9th Python in Science Conference.SciPy;2010;92-96.Accessed August 8, 2023.https://conference.scipy.org/proceedings/scipy2010/seabold.html
-
13 ↑
KlineA.PsmPy: a package for retrospective cohort matching in Python. In:44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC).IEEE;2022:1354-1357.Accessed August 8, 2023.https://ieeexplore.ieee.org/document/9871333
-
14 ↑
BarzóP,MarmarouA,FatourosP,CorwinF,DunbarJ.Magnetic resonance imaging-monitored acute blood-brain barrier changes in experimental traumatic brain injury.J Neurosurg.1996;85(6):1113–1121.
-
15
CzignerA,MihályA,FarkasO,et al.Kinetics of the cellular immune response following closed head injury.Acta Neurochir (Wien).2007;149(3):281–289.
-
16 ↑
LokJ,LeungW,MurphyS,ButlerW,NoviskiN,LoEH.Intracranial hemorrhage: mechanisms of secondary brain injury.Acta Neurochir Suppl.2011;111(63):69.
-
17 ↑
CarneyN,TottenAM,O’ReillyC,et al.Guidelines for the Management of Severe Traumatic Brain Injury,.Fourth Edition.开云体育app官方网站下载入口.2017;80(1):6–15.
-
18 ↑
PowersWJ,RabinsteinAA,AckersonT,et al.2018Guidelines for the Early Management of Patients With Acute Ischemic Stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association.Stroke.2018;49(3):e46–e110.
-
19
AlerhandS,LayC.Spontaneous intracerebral hemorrhage.Emerg Med Clin North Am.2017;35(4):825–845.
-
20 ↑
GrossBA,JankowitzBT,FriedlanderRM.Cerebral intraparenchymal hemorrhage: a review.JAMA.2019;321(13):1295–1303.
-
21 ↑
LiuZH,ChenNY,TuPH,et al.Previous antithrombotic therapy, particularly anticoagulant, is associated with unfavorable outcomes in patients with primary spontaneous intracerebral hemorrhage receiving craniotomy: a nationwide population-based cohort study.World Neurosurg.2019;128:e59–e73.
-
22 ↑
BaraL,BillaudE,GramondG,KherA,SamamaM.Comparative pharmacokinetics of a low molecular weight heparin (PK 10 169) and unfractionated heparin after intravenous and subcutaneous administration.Thromb Res.1985;39(5):631–636.
-
23 ↑
ConstantiniS,KannerA,FriedmanA,et al.Safety of perioperative minidose heparin in patients undergoing brain tumor surgery: a prospective, randomized, double-blind study.J Neurosurg.2001;94(6):918–921.
-
24
AltshulerE,MoosaH,SelkerRG,VertosickFTJr.The risk and efficacy of anticoagulant therapy in the treatment of thromboembolic complications in patients with primary malignant brain tumors.开云体育app官方网站下载入口.1990;27(1):74–76.
-
25
TykockiT,GuzekK.Anticoagulation therapy in traumatic brain injury.World Neurosurg.2016;89:497–504.
-
26 ↑
HantscheA,WilhelmyF,KasperJ,et al.Early prophylactic anticoagulation after subarachnoid hemorrhage decreases systemic ischemia and improves outcome.Clin Neurol Neurosurg.2021;207:106809.
-
27 ↑
DickinsonLD,MillerLD,PatelCP,GuptaSK.Enoxaparin increases the incidence of postoperative intracranial hemorrhage when initiated preoperatively for deep venous thrombosis prophylaxis in patients with brain tumors.开云体育app官方网站下载入口.1998;43(5):1074–1081.
-
28 ↑
ZakaryanA.Perioperative management of neurosurgical patients receiving chronic anticoagulation therapy.Front Pharmacol.2014;5:64.
-
29 ↑
Parry-JonesAR,MoullaaliTJ,ZiaiWC.Treatment of intracerebral hemorrhage: from specific interventions to bundles of care.Int J Stroke.2020;15(9):945–953.
-
30 ↑
SoW,HugenholtzH,RichardMT.Complications of anticoagulant therapy in patients with known central nervous system lesions.Can J Surg.1983;26(2):181–183.
-
31 ↑
PriorA,FiaschiP,IaccarinoC,et al.How do you manage ANTICOagulant therapy in neurosurgery? The ANTICO survey of the Italian Society of Neurosurgery (SINCH).BMC Neurol.2021;21(1):98.
-
32 ↑
DouketisJD,SpyropoulosAC,SpencerFA,et al.Perioperative management of antithrombotic therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines.Chest.2012;141(2 Suppl):e326S–e350S.
-
33 ↑
BadireddyM,MudipalliVR.Deep venous thrombosis prophylaxis. In:StatPearls.StatPearls Publishing;2023.Accessed April 21, 2023.http://www.ncbi.nlm.nih.gov/books/NBK534865/
-
34 ↑
AllmanRM,GoodePS,BurstN,BartolucciAA,ThomasDR.Pressure ulcers, hospital complications, and disease severity: impact on hospital costs and length of stay.Adv Wound Care.1999;12(1):22–30.
-
35
MastanehZ,MouseliA,MohseniS,DadipoorS.Predictors of hospital length of stay and mortality among COVID-19 inpatients during 2020-2021 in Hormozgan Province of Iran: a retrospective cohort study.Health Sci Rep.2023;6(6):e1329.
-
36
KushnerDS,Johnson-GreeneD,FelixER,MillerC,CorderoMK,ThomashawSA.Predictors of discharge to home/community following inpatient-rehabilitation in a US national sample of Guillain-Barre-Syndrome patients.PLoS One.2023;18(5):e0286296.
-
37 ↑
DavidsonIU,BrigatiDP,FaourM,UdoInyangIJ,IbrahimM,MurrayTG.Same-day bilateral total knee arthroplasty candidacy criteria decrease length of stay and facility discharge.Orthopedics.2018;41(5):293–298.
-
38 ↑
WooD,ComeauME,VenemaSU,et al.Risk factors associated with mortality and neurologic disability after intracerebral hemorrhage in a racially and ethnically diverse cohort.JAMA Netw Open.2022;5(3):e221103.
-
39
KuohnLR,LeasureAC,AcostaJN,et al.Cause of death in spontaneous intracerebral hemorrhage survivors: multistate longitudinal study.Neurology.2020;95(20):e2736–e2745.
-
40
FogelholmR,MurrosK,RissanenA,AvikainenS.Long term survival after primary intracerebral haemorrhage: a retrospective population based study.J Neurol Neurosurg Psychiatry.2005;76(11):1534–1538.
-
41 ↑
YueJK,KrishnanN,KanterJH,et al.Neuroworsening in the emergency department is a predictor of traumatic brain injury intervention and outcome: a TRACK-TBI pilot study.J Clin Med.2023;12(5):2024.
-
42 ↑
GreuterL,UllmannM,GuzmanR,SolemanJ.Mortality of surgically treated neurotrauma in elderly patients and the development of a prediction score: Geriatric Neurotrauma Mortality Score.World Neurosurg.2023;175:e1–e20.
-
43 ↑
GuermaziA,MiauxY,Rovira-CañellasA,et al.Neuroradiological findings in vascular dementia.Neuroradiology.2007;49(1):1–22.
