Abstract
Background
Lumbar punctures are one of the most used diagnostic and therapeutic procedures for a range of CNS pathologies, including infection and increased intracranial pressure due to many etiologies. Lumbar punctures violate the dura of the thecal sac, posing some risk for the formation of epidural or subdural hematomas. This risk may be increased in patients on anticoagulant or antiplatelet therapy. This study compares the odds of developing epidural or subdural hematomas, paralysis, and the requirement of epidural blood for hemostasis within 30 days of lumbar puncture in patients with and without anticoagulant or antiplatelet use in hopes of establishing evidence-based guidelines for clinical decision-making for lumbar punctures in patients using these medications.
Methods
The TriNetX multi-institutional electronic health record database was used to perform a retrospective, propensity score-matched analysis of outcomes of cohorts of patients who underwent lumbar punctures on anticoagulants, antiplatelets, and no anticoagulants or antiplatelets within one month of the procedure. The outcomes of interest were examined within 30 days of lumbar puncture and included the occurrence of epidural or subdural hematoma, paralysis, requirement of an epidural blood patch, and requirement of spinal decompression.
Results
A total of 25,788 patients were identified for both lumbar puncture patients with anticoagulants and without blood thinners after propensity score matching. For this study, epidural hematoma (OR [95%CI], 1.000 [0.561–1.784]) showed no difference in odds ratio between the two cohorts, while subdural hematoma (OR [95%CI], 0.780 [0.640–0.950]) showed a statistically significantly lower odds ratio for the anticoagulant use group. We identified 13,310 patients for both lumbar puncture patients with antiplatelets and lumbar puncture patients without antiplatelet, anticoagulant, heparin, and LMWH (“blood thinners”) after propensity score matching. For this analysis, epidural hematoma (OR [95%CI], 1.904 [0.885–4.095]) showed a higher odds ratio compared to subdural hematoma (OR [95%CI], 1.008 [0.784–1.296]), both of which were not statistically significant.
Conclusion
The data analyses suggest that there are no increased risks of epidural and subdural hematoma in patients with a history of use of antiplatelet or anticoagulant compared to those who do not. These results could be further validated by a single or multi-institutional retrospective cohort study with additional parameters that characterize the patient population.
Keywords: Subdural, Epidural, Hematoma, Lumbar puncture, Anticoagulant, Antiplatelet
Introduction
Lumbar puncture is a commonly used invasive test used to diagnose and treat headaches, inflammatory diseases, and infectious diseases of the nervous system [1]. Diagnostics using lumbar punctures involve collecting cerebrospinal fluid samples for lab testing [2]. An example of an infectious disease of the nervous system whereby lumbar puncture diagnostics are essential is meningitis [3]. Lumbar puncture procedures are relatively safe as complications rarely arise [4]. Although relatively rare, lumbar puncture procedures are associated with complications such as spinal hematomas, pain at the puncture site, dizziness, and post-lumbar headache (PLPH). The most common complication observed is PLPH, with an incidence of 32% [1]. Spinal hematoma, which is an accumulation of blood in the spinal canal, is among the most critical complications associated with lumbar puncture procedures [3, 5]. Spinal hematomas typically lead to a compression of the spinal cord, leading to various neurological deficits [6]. A widely used remedy for that is spinal decompression, which is a conservative treatment method for intervertebral disc herniation. Spinal decompression works by reducing pressure on the spinal cord by supplying nutrients and oxygen to the disc [7]. There exist three main types of spinal hematomas: subdural, epidural, and subarachnoid hematomas [6]. A spinal subdural hematoma, also called spinal subdural hemorrhage, is an accumulation of blood in subdural space, which is the space between the spinal arachnoid and the dura membrane [8]. A spinal epidural hematoma is a collection of blood between the spinal dura mater and the surface of the vertebrae. Spinal hematomas are relatively rare neurosurgical emergencies, and when they occur, they are often fatal, with a death rate of about 50% [9].
Anticoagulants are widely used drugs that prevent blood coagulation or clotting. They are mostly used in venous thromboembolism (VTE) treatment [10]. Antiplatelets, on the other hand, help prevent the accumulation of platelets, which is an important pathophysiological aspect of vascular accidents such as hematomas [11]. Understanding the dearth of literature in the realm of understanding how the history of use of anticoagulants or antiplatelets affects the occurrence of subdural and epidural hematomas, this study investigates the risk of developing subdural and epidural hematoma following lumbar puncture for patients with a history of antiplatelet and anticoagulant use. We hypothesize that a history of use of both anticoagulants and antiplatelets would increase the risks of developing subdural and epidural hematomas [12].
Methods
Study design
This retrospective cohort study used the TriNetX Research Network database, which includes data from over 106 million patients from 68 healthcare networks across the world, to query for patients who underwent lumbar punctures with and without anticoagulant and antiplatelet use using the International Classification of Diseases, 10th Revision (ICD-10), current procedural terminology (CPT) codes, RxNorm, and VA formulary medication systems. TriNetX is a de-identified patient database that uses electronic medical records from multiple institutions worldwide. No Institutional Review Board (IRB) approval or patient consent was required due to the de-identified nature of the database. Previous literature confirmed the utility and validity of this database, and its details are well-described [13–15].
Patient selection
The TriNetX database we queried on September 25, 2022, for patients of any age and included outcomes from LPs with and without anticoagulant or antiplatelet use that occurred within the past 20 years. Patients that underwent lumbar punctures were selected using CPT codes 00635, 62,270, 1,035,653, and 62,328 and ICD-10 codes 009U3ZX, 009U3ZZ, 009U30Z, 009U4ZX, 009U40Z, and 009U4ZZ. A control cohort included any patients who underwent any of the aforementioned LP procedures and were not exposed to anticoagulants or antiplatelets. Patients in the anticoagulant-receiving cohort were further classified as having been exposed to enoxaparin, heparin, and other anticoagulants using RxNorm codes 67,108 and 5224, along with VA formulary code BL110, respectively. Patients in the antiplatelet cohort were selected using RxNorm codes 32,968, 10,594, 1191, 1,116,632, 3521, and 613,391 for exposure to clopidogrel, aspirin, prasugrel, and more. Patients who received protamine sulfate, idarucizumab, platelet transfusions, prothrombin, blood, or blood components were excluded due to the potential reversal of anticoagulation.
Using CPT codes, the resulting patient cohorts were examined for covariates that may confound results, including thrombocytopenia, Von Willebrand’s disease, platelet defects, age, sex, demographics, and more. The cohorts were then propensity score-matched in an attempt to mimic true randomization.
Statistical analysis
Various statistical tools, including risk difference (formerly referred to as attributable risk), odds ratio (OR), risk ratio (RR, also known as relative risk), and standard mean difference (SMD), were used to analyze data. Risk difference is helpful for quantifying the difference in risk of experiencing an outcome between exposed and unexposed cohorts. The risk ratio quantifies the ratio of the risk of experiencing an outcome between exposed and unexposed cohorts. The odds ratio quantifies the odds of experiencing an outcome between exposed and unexposed cohorts [16]. SMD quantifies the difference between cohorts, where a value of zero indicates no difference between cohorts [17]. An alpha level of 0.05 was used to determine statistical significance. Data was analyzed using TriNetX software with JAVA, R, and Python programming languages.
Outcomes were examined within 30 days of LP and included the occurrence of epidural or subdural hematoma (EDH and SDH, respectively), paralysis, requirement of an epidural blood patch (EBP), and requirement of spinal decompression. The 30-day time-limit excludes complications unrelated to LP.
Results
Before propensity score-matching, we identified 26,704 lumbar puncture patients with anticoagulant use history and 165,605 lumbar puncture patients without antiplatelet, anticoagulant, heparin, and LMWH (“blood thinners”) use. Age at index were 48.8 ± 22.2 years and 28.5 ± 23.2 years for anticoagulant-exposed and unexposed cohorts, respectively. 47.8% of the anticoagulant-exposed cohort and 44.5% of non-anticoagulant-exposed cohort were males.
Pre-propensity score-matching baseline characteristics and demographics are shown in Table 1 for the anticoagulant versus non-anticoagulant comparative study.
Table 1.
Pre-propensity score-matching baseline characteristics and demographics of lumbar puncture patients with and without anticoagulant use history
| Baseline characteristic | Anticoagulant-exposed | % of cohort | Non-anticoagulant-exposed | % of Cohort | p-value | Standard mean difference |
|---|---|---|---|---|---|---|
| Patients | 26,704 | 100% | 165,605 | 100% | ||
| Age at Index | 48.8 (+/- 22.2 SD) | 100% | 28.5 (+/- 23.2 SD) | 100% | < 0.001 | 0.894 |
| Male | 12,756 | 47.8% | 66,381 | 44.5% | < 0.001 | 0.066 |
| Female | 13,905 | 52.2% | 82,666 | 55.5% | < 0.001 | 0.066 |
| White | 16,756 | 62.8% | 92,751 | 62.2% | 0.053 | 0.013 |
| Black or African American | 5,356 | 20.1% | 25,216 | 16.9% | < 0.001 | 0.082 |
| Asian | 567 | 2.1% | 2,798 | 1.9% | 0.006 | 0.018 |
| Unknown Race | 3,771 | 14.1% | 27,230 | 18.3% | < 0.001 | 0.112 |
| Fibrosis and Cirrhosis of Liver | 533 | 2.0% | 945 | 0.6% | < 0.001 | 0.120 |
| Hypertensive Diseases | 10,741 | 40.3% | 20,179 | 13.5% | < 0.001 | 0.633 |
| Diabetes Mellitus | 4,958 | 18.6% | 8,065 | 5.4% | < 0.001 | 0.414 |
| Overweight, Obesity, Other Hyperalimentation | 3,936 | 14.8% | 10,279 | 6.9% | < 0.001 | 0.255 |
| Nicotine Dependence | 4,322 | 16.2% | 11,080 | 7.4% | < 0.001 | 0.274 |
| Alcohol Abuse | 1,229 | 4.6% | 2,131 | 1.4% | < 0.001 | 0.187 |
| Alcohol Dependence | 1,184 | 4.4% | 1,745 | 1.2% | < 0.001 | 0.199 |
| Chronic Lower Respiratory Diseases | 4,145 | 15.5% | 12,992 | 8.7% | < 0.001 | 0.210 |
| Atrial Fibrillation and Atrial Flutter | 1,810 | 6.8% | 1,497 | 1.0% | < 0.001 | 0.302 |
| Heart Failure | 2,011 | 7.5% | 2,129 | 1.4% | < 0.001 | 0.299 |
| Ischemic Heart Diseases | 3,463 | 13.0% | 3,176 | 2.1% | < 0.001 | 0.420 |
| Other Peripheral Vascular Diseases | 906 | 3.4% | 1,265 | 0.8% | < 0.001 | 0.178 |
| Thrombocytopenia, Unspecified | 1,906 | 7.1% | 3,581 | 2.4% | < 0.001 | 0.224 |
| Immune Thrombocytopenic Purpura | 69 | 0.3% | 186 | 0.1% | < 0.001 | 0.031 |
| Qualitative Platelet Defects | 50 | 0.2% | 93 | 0.1% | < 0.001 | 0.035 |
| Von Willebrand’s Disease | 19 | 0.1% | 97 | 0.1% | 0.717 | 0.002 |
| Hereditary Factor XI Deficiency | 10 | 0.0% | 12 | 0.0% | < 0.001 | 0.020 |
| Disseminated Intravascular Coagulation | 141 | 0.5% | 271 | 0.2% | < 0.001 | 0.058 |
| Hereditary Factor VIII Deficiency | 10 | 0.0% | 37 | 0.0% | 0.243 | 0.007 |
| Hereditary Factor IX Deficiency | 10 | 0.0% | 15 | 0.0% | 0.001 | 0.018 |
| Activated Protein C Resistance | 107 | 0.4% | 140 | 0.1% | < 0.001 | 0.062 |
| Prothrombin Gene Mutation | 87 | 0.3% | 121 | 0.1% | < 0.001 | 0.054 |
| Antiphospholipid Syndrome | 107 | 0.4% | 135 | 0.1% | < 0.001 | 0.063 |
| Lupus Anticoagulant Syndrome | 97 | 0.4% | 131 | 0.1% | < 0.001 | 0.058 |
This table displays the baseline characteristics and demographics of lumbar puncture patients in the anticoagulant-exposed and unexposed cohorts before propensity score matching. An alpha level of 0.05 was selected for statistical significance
SD: standard deviation
We identified 14,395 antiplatelet-exposed lumbar puncture patients and 165,605 non-antiplatelet-exposed (including antiplatelet, anticoagulant, heparin, and LMWH (“blood thinners”)) lumbar puncture patients before propensity score-matching. Age at index were 59.2 ± 17.8 years and 28.5 ± 23.2 years for antiplatelet-exposed and unexposed cohorts, respectively. 49.8% of the antiplatelet-exposed cohort and 44.5% of the non-antiplatelet-exposed cohort were males. Pre-propensity score-matching baseline characteristics and demographics are shown in Table 2 for the antiplatelet versus non-antiplatelet comparative study.
Table 2.
Pre-propensity score-matching baseline characteristics and demographics of lumbar puncture patients with and without antiplatelet use history
| Baseline characteristic | Antiplatelet-exposed | % of cohort | Non-antiplatelet-exposed | % of cohort | p-value | Standard mean difference |
|---|---|---|---|---|---|---|
| Patients | 14,395 | 100% | 165,605 | 100% | ||
| Age at Index | 59.2 (+/- 17.8 SD) | 100% | 28.5 (+/- 23.2 SD) | 100% | < 0.001 | 1.485 |
| Male | 7,117 | 49.8% | 66,381 | 44.5% | < 0.001 | 0.105 |
| Female | 7,176 | 50.2% | 82,666 | 55.5% | < 0.001 | 0.105 |
| White | 9,413 | 65.9% | 92,751 | 62.2% | < 0.001 | 0.076 |
| Black or African American | 2,758 | 19.3% | 25,216 | 16.9% | < 0.001 | 0.062 |
| Asian | 234 | 1.6% | 2,798 | 1.9% | 0.042 | 0.018 |
| Unknown Race | 1,774 | 12.4% | 27,230 | 18.3% | < 0.001 | 0.163 |
| Fibrosis and Cirrhosis of Liver | 365 | 2.6% | 945 | 0.6% | < 0.001 | 0.154 |
| Hypertensive Diseases | 8,048 | 56.3% | 20,179 | 13.5% | < 0.001 | 1.004 |
| Diabetes Mellitus | 4,076 | 28.5% | 8,065 | 5.4% | < 0.001 | 0.647 |
| Overweight, Obesity, Other Hyperalimentation | 2,630 | 18.4% | 10,279 | 6.9% | < 0.001 | 0.351 |
| Nicotine Dependence | 2,323 | 16.3% | 11,080 | 7.4% | < 0.001 | 0.276 |
| Alcohol Abuse | 680 | 4.8% | 2,131 | 1.4% | < 0.001 | 0.193 |
| Alcohol Dependence | 624 | 4.4% | 1,745 | 1.2% | < 0.001 | 0.196 |
| Chronic Lower Respiratory Diseases | 2,791 | 19.5% | 12,992 | 8.7% | < 0.001 | 0.314 |
| Atrial Fibrillation and Atrial Flutter | 1,804 | 12.6% | 1,497 | 1.0% | < 0.001 | 0.474 |
| Heart Failure | 2,075 | 14.5% | 2,129 | 1.4% | < 0.001 | 0.498 |
| Ischemic Heart Diseases | 3,556 | 24.9% | 3,176 | 2.1% | < 0.001 | 0.706 |
| Other Peripheral Vascular Diseases | 882 | 6.2% | 1,265 | 0.8% | < 0.001 | 0.292 |
| Thrombocytopenia, Unspecified | 1,326 | 9.3% | 3,581 | 2.4% | < 0.001 | 0.296 |
| Immune Thrombocytopenic Purpura | 47 | 0.3% | 186 | 0.1% | < 0.001 | 0.043 |
| Qualitative Platelet Defects | 49 | 0.3% | 93 | 0.1% | < 0.001 | 0.062 |
| Von Willebrand’s Disease | 10 | 0.1% | 97 | 0.1% | 0.827 | 0.002 |
| Hereditary Factor XI Deficiency | 10 | 0.1% | 12 | 0.0% | < 0.001 | 0.031 |
| Disseminated Intravascular Coagulation | 100 | 0.7% | 271 | 0.2% | < 0.001 | 0.078 |
| Hereditary Factor VIII Deficiency | 10 | 0.1% | 37 | 0.0% | 0.002 | 0.021 |
| Hereditary Factor IX Deficiency | 10 | 0.1% | 15 | 0.0% | < 0.001 | 0.030 |
| Activated Protein C Resistance | 99 | 0.7% | 140 | 0.1% | < 0.001 | 0.096 |
| Prothrombin Gene Mutation | 83 | 0.6% | 121 | 0.1% | < 0.001 | 0.087 |
| Antiphospholipid Syndrome | 102 | 0.7% | 135 | 0.1% | < 0.001 | 0.099 |
| Lupus Anticoagulant Syndrome | 95 | 0.7% | 131 | 0.1% | < 0.001 | 0.094 |
This table displays the baseline characteristics and demographics of the patients in the antiplatelet-exposed and unexposed cohorts before propensity score matching. An alpha level of 0.05 was selected for statistical significance
SD: standard deviation
After propensity score-matching, 25,788 patients were identified for both lumbar puncture patient cohorts with and without anticoagulant exposure. Age at index were 47.9 ± 22.0 years and 48.2 ± 22.1 years, respectively. 47.3% and 47.4% were males in the anticoagulant-use and non-anticoagulant-use cohorts, respectively. Post-propensity score-matching baseline characteristics and demographics are shown in Table 3 for the anticoagulant versus non-anticoagulant comparative study. We identified 13,310 lumbar puncture patients for both antiplatelet-exposed and non-antiplatelet-exposed cohorts after propensity score-matching, and their age at index were 58.0 ± 17.8 years and 59.0 ± 18.0 years, respectively. 49.0% of the antiplatelet-exposed cohort and 48.5% of the non-antiplatelet-exposed cohort were males. Post-propensity score-matching baseline characteristics and demographics are shown in Table 4 for the antiplatelet versus non-antiplatelet comparative study.
Table 3.
Post-propensity score-matching baseline characteristics and demographics of lumbar puncture patients with and without anticoagulant use history
| Baseline characteristic | Anticoagulant-exposed | % of cohort | Non-anticoagulant-exposed | % of cohort | p-value | Standard mean difference |
|---|---|---|---|---|---|---|
| Patients | 25,788 | 100% | 25,788 | 100% | ||
| Age at Index | 47.9 (+/- 22.0 SD) | 100% | 48.2 (+/- 22.1 SD) | 100% | ||
| Male | 12,209 | 47.3% | 12,216 | 47.4% | 0.951 | 0.001 |
| Female | 13,577 | 52.6% | 13,567 | 52.6% | 0.930 | 0.001 |
| White | 16,179 | 62.7% | 16,165 | 62.7% | 0.899 | 0.001 |
| Black or African American | 5,173 | 20.1% | 5,444 | 21.1% | 0.003 | 0.026 |
| Asian | 542 | 2.1% | 545 | 2.1% | 0.927 | 0.001 |
| Unknown Race | 3,691 | 14.3% | 3,441 | 13.3% | 0.001 | 0.028 |
| Fibrosis and Cirrhosis of Liver | 493 | 1.9% | 443 | 1.7% | 0.099 | 0.015 |
| Hypertensive Diseases | 9,897 | 38.4% | 9,945 | 38.6% | 0.664 | 0.004 |
| Diabetes Mellitus | 4,432 | 17.2% | 4,420 | 17.1% | 0.889 | 0.001 |
| Overweight, Obesity, Other Hyperalimentation | 3,562 | 13.8% | 3,473 | 13.5% | 0.254 | 0.010 |
| Nicotine Dependence | 4,007 | 15.5% | 4,096 | 15.9% | 0.282 | 0.009 |
| Alcohol Abuse | 1,105 | 4.3% | 1,094 | 4.2% | 0.811 | 0.002 |
| Alcohol Dependence | 1,046 | 4.1% | 966 | 3.7% | 0.069 | 0.016 |
| Chronic Lower Respiratory Diseases | 3,772 | 14.6% | 3,653 | 14.2% | 0.136 | 0.013 |
| Atrial Fibrillation and Atrial Flutter | 1,364 | 5.3% | 1,196 | 4.6% | 0.001 | 0.030 |
| Heart Failure | 1,630 | 6.3% | 1,442 | 5.6% | < 0.001 | 0.031 |
| Ischemic Heart Diseases | 2,766 | 10.7% | 2,518 | 9.8% | < 0.001 | 0.032 |
| Other Peripheral Vascular Diseases | 739 | 2.9% | 632 | 2.5% | 0.003 | 0.026 |
| Thrombocytopenia, Unspecified | 1,650 | 6.4% | 1,625 | 6.3% | 0.652 | 0.004 |
| Immune Thrombocytopenic Purpura | 59 | 0.2% | 51 | 0.2% | 0.445 | 0.007 |
| Qualitative Platelet Defects | 40 | 0.2% | 32 | 0.1% | 0.345 | 0.008 |
| Von Willebrand’s Disease | 19 | 0.1% | 19 | 0.1% | 1.00 | < 0.001 |
| Hereditary Factor XI Deficiency | 10 | 0.0% | 10 | 0.0% | ‘1.00 | < 0.001 |
| Disseminated Intravascular Coagulation | 123 | 0.5% | 110 | 0.4% | 0.393 | 0.008 |
| Hereditary Factor VIII Deficiency | 10 | 0.0% | 10 | 0.0% | ‘1.00 | < 0.001 |
| Hereditary Factor IX Deficiency | 10 | 0.0% | 10 | 0.0% | ‘1.00 | < 0.001 |
| Activated Protein C Resistance | 89 | 0.3% | 79 | 0.3% | 0.440 | 0.007 |
| Prothrombin Gene Mutation | 74 | 0.3% | 65 | 0.3% | 0.445 | 0.007 |
| Antiphospholipid Syndrome | 90 | 0.3% | 75 | 0.3% | 0.242 | 0.010 |
| Lupus Anticoagulant Syndrome | 81 | 0.3% | 72 | 0.3% | 0.466 | 0.006 |
This table displays the baseline characteristics and demographics of the patients in the anticoagulant-exposed and unexposed cohorts after propensity score matching. An alpha level of 0.05 was selected for statistical significance
SD: standard deviation
Table 4.
Post-propensity score-matching baseline characteristics and demographics of lumbar puncture patients with and without antiplatelet use history
| Baseline characteristic | Antiplatelet-exposed | % of cohort | Non-antiplatelet-exposed | % of cohort | p-Value | Standard mean difference |
|---|---|---|---|---|---|---|
| Patients | 13,310 | 100% | 13,310 | 100% | ||
| Age at Index | 58.0 (+/- 17.8 SD) | 100% | 59.0 (+/- 18.0 SD) | 100% | < 0.001 | 0.056 |
| Male | 6,520 | 49.0% | 6,450 | 48.5% | 0.391 | 0.011 |
| Female | 6,789 | 51.0% | 6,855 | 51.5% | 0.418 | 0.01 |
| White | 8,721 | 65.5% | 8,631 | 64.8% | 0.247 | 0.014 |
| Black or African American | 2,577 | 19.4% | 2,689 | 20.2% | 0.085 | 0.021 |
| Asian | 226 | 1.7% | 268 | 2.0% | 0.056 | 0.023 |
| Unknown Race | 1,679 | 12.6% | 1,624 | 12.2% | 0.307 | 0.013 |
| Fibrosis and Cirrhosis of Liver | 320 | 2.4% | 305 | 2.3% | 0.544 | 0.007 |
| Hypertensive Diseases | 7,088 | 53.3% | 7,137 | 53.6% | 0.547 | 0.007 |
| Diabetes Mellitus | 3,449 | 25.9% | 3,396 | 25.5% | 0.457 | 0.009 |
| Overweight, Obesity, Other Hyperalimentation | 2,225 | 16.7% | 2,045 | 15.4% | 0.003 | 0.037 |
| Nicotine Dependence | 2,100 | 15.8% | 2,004 | 15.1% | 0.103 | 0.020 |
| Alcohol Abuse | 614 | 4.6% | 583 | 4.4% | 0.359 | 0.011 |
| Alcohol Dependence | 543 | 4.1% | 543 | 4.1% | 1.00 | < 0.001 |
| Chronic Lower Respiratory Diseases | 2,385 | 17.9% | 2,249 | 16.9% | 0.028 | 0.027 |
| Atrial Fibrillation and Atrial Flutter | 1,256 | 9.4% | 1,065 | 8.0% | < 0.001 | 0.051 |
| Heart Failure | 1,510 | 11.3% | 1,315 | 9.9% | < 0.001 | 0.048 |
| Ischemic Heart Diseases | 2,685 | 20.2% | 2,333 | 17.5% | < 0.001 | 0.068 |
| Other Peripheral Vascular Diseases | 652 | 4.9% | 561 | 4.2% | 0.007 | 0.033 |
| Thrombocytopenia, Unspecified | 1,072 | 8.1% | 988 | 7.4% | 0.054 | 0.024 |
| Immune Thrombocytopenic Purpura | 39 | 0.3% | 38 | 0.3% | 0.909 | 0.001 |
| Qualitative Platelet Defects | 33 | 0.2% | 25 | 0.2% | 0.293 | 0.013 |
| Von Willebrand’s Disease | 10 | 0.1% | 10 | 0.1% | 1.00 | < 0.001 |
| Hereditary Factor XI Deficiency | 10 | 0.1% | 0 | 0% | 0.002 | 0.039 |
| Disseminated Intravascular Coagulation | 81 | 0.6% | 68 | 0.5% | 0.286 | 0.013 |
| Hereditary Factor VIII Deficiency | 10 | 0.1% | 10 | 0.1% | 1.00 | < 0.001 |
| Hereditary Factor IX Deficiency | 10 | 0.1% | 10 | 0.1% | 1.00 | < 0.001 |
| Activated Protein C Resistance | 70 | 0.5% | 58 | 0.4% | 0.288 | 0.013 |
| Prothrombin Gene Mutation | 60 | 0.5% | 49 | 0.4% | 0.291 | 0.013 |
| Antiphospholipid Syndrome | 73 | 0.5% | 59 | 0.4% | 0.222 | 0.015 |
| Lupus Anticoagulant Syndrome | 66 | 0.5% | 53 | 0.4% | 0.232 | 0.015 |
This table displays the baseline characteristics and demographics of the patients in the antiplatelet-exposed and unexposed cohorts after propensity score matching. An alpha level of 0.05 was selected for statistical significance
SD: standard deviation
Table 5 describes the outcomes of subdural and epidural hematomas that occurred within 30 days after undergoing lumbar puncture by comparing the cohorts with and without a history of anticoagulant use. When comparing the occurrence of adverse outcomes for the anticoagulant-exposed cohort to the non-anticoagulant-exposed cohort, epidural hematoma (OR [95%CI], 1.000 [0.561–1.784]) showed no difference in odds ratio between two cohorts with no statistical significance. However, subdural hematoma (OR [95%CI], 0.780 [0.640–0.950]) showed a statistically significantly lower odds ratio for the anticoagulant use group. Table 6 shows the outcomes of subdural and epidural hematomas that occurred within 30 days after undergoing lumbar puncture by comparing the cohorts with and without a history of antiplatelet use. Both subdural hematoma and epidural hematoma outcomes were not statistically significantly increased in one cohort versus another.
Table 5.
Risk differences, risk ratios, and odds ratios of developing subdural hematoma and epidural hematoma within 30 days after lumbar puncture for patients with and without anticoagulant use history
| Outcome | Risk difference | 95% CI | p-Value | Risk ratio | 95% CI | Odds ratio | 95% CI |
|---|---|---|---|---|---|---|---|
| Subdural Hematoma | -0.002 | (-0.004, -0.000) | 0.013 | 0.782 | (0.643, 0.951) | 0.78 | (0.640, 0.950) |
| Epidural Hematoma | 0.000 | (-0.001, 0.001) | 0.999 | 1.000 | (0.561, 1.783) | 1.000 | (0.561, 1.784) |
This table displays the risk differences, risk ratios, and odds ratios comparing various outcomes experienced within 30 days of lumbar puncture by the anticoagulant exposed and unexposed cohorts. An alpha level of 0.05 was selected for statistical significance
CI: confidence interval
Table 6.
Risk differences, risk ratios, and odds ratios of developing subdural hematoma and epidural hematoma within 30 days after lumbar puncture for patients with and without antiplatelet use history
| Outcome | Risk difference | 95% CI | p-Value | Risk Ratio | 95% CI | Odds ratio | 95% CI |
|---|---|---|---|---|---|---|---|
| Subdural Hematoma | 0.000 | (-0.002, 0.002) | 0.952 | 1.008 | (0.785, 1.293) | 1.008 | (0.784, 1.296) |
| Epidural Hematoma | 0.001 | (-0.000, 0.001 | 0.094 | 1.902 | (0.885, 4.089) | 1.904 | (0.885, 4.095) |
This table displays the risk differences, risk ratios, and odds ratios comparing various outcomes experienced within 30 days of lumbar puncture by the antiplatelet exposed and unexposed cohorts. An alpha level of 0.05 was selected for statistical significance
CI: confidence interval
Discussion
The present study evaluated the risk of developing subdural hematoma and epidural hematoma after a lumbar puncture in patients with a history of antiplatelet and anticoagulation therapy. Many case reports have shown examples of complications, such as hematomas and bleeding following lumbar punctures, with patients on antiplatelet and anticoagulation therapy. In contrast to these studies, this study offers a multi-institutional view of the risk associated with lumbar puncture while on antiplatelet or anticoagulation therapy. Utilizing a large cohort population, a clearer association can be established between anticoagulation and antiplatelet medication use and the risk of complications associated with lumbar puncture.
Subdural hematoma and antiplatelets
Based on our results, there is no increased risk of subdural hematoma following lumbar puncture in patients with antiplatelet therapy (OR [95%CI], 1.008 [0.784–1.296]). Similar findings were shown in a prospective cohort study conducted by Horlocker et al., where 924 patients underwent spinal anesthesia on antiplatelet therapy. The study found that minor bleeding complications were increased on antiplatelet therapy, but none of the patients developed spinal hematomas [18]. Additionally, a literature review spanning from 1906 to 1994 conducted by Vandermeulen et al. evaluated bleeding risk following lumbar punctures and found that only 3 out of 61 cases of bleeding occurred in patients with antiplatelet use [19]. While its antiplatelet therapy, such as aspirin, may not increase the risk of hematomas, combination with anticoagulation as well as dual antiplatelet therapy has shown an increased risk [20–22]. This may be of interest for future investigation involving large multi-institutional representative population-level data. Overall, most literature is concurrent with the current study finding that antiplatelet therapy does not significantly increase the risk of subdural hematoma following lumbar puncture.
Subdural hematoma and anticoagulants
However, in contrast to the existing literature, our study showed a slightly decreased risk of developing a subdural hematoma for anticoagulation use cohort (OR [95%CI], 0.780 [0.640–0.950]). This finding can be best explained by analyzing the existing evidence regarding subdural hematomas following lumbar punctures. In patients with anticoagulation therapy, an increased risk of hematomas and other major bleeding complications has been shown by various studies such as the one by Ruff et al. [20] However, this study initiated anticoagulation after the lumbar puncture and did not explicitly study the risk of hematomas in patients already on anticoagulation therapy, which was the focus of our study. Many case studies and case series have also shown an increased risk of subdural hematoma in patients with anticoagulation [23–25]. However, these studies, specifically those conducted by Ito et al. and Kosarchuk et al., do not include procedural or iatrogenic causes of subdural hematoma but rather spontaneous cases [23, 25]. As a result, comparing our study’s results directly with the referenced study models and criteria poses challenges, and our study, therefore, adds value and a new perspective to the current existing literature.
Epidural hematoma and antiplatelets
From our study, we found that there was no statistically significant association between the use of antiplatelet therapy and risk for epidural hematoma following lumbar puncture (OR [95%CI], 1.904 [0.885–4.095]). This finding is consistent with two retrospective studies analyzing epidural hematoma formation following spinal anesthesia, an equivalently invasive procedure to lumbar puncture. A study conducted by Horlocker et al. reported that out of a total of 391 patients on preoperative antiplatelet medication, none developed signs of spinal hematoma [26]. The impact of other antiplatelet medications, including thienopyridines such as clopidogrel, ticlopidine, and prasugrel, is largely unclear as only small-sampled case studies show an increased risk of hematoma formation after lumbar puncture [27, 28]. However, our study at a larger scale includes these antiplatelet medications in its analysis, demonstrating that the overall risk of spinal hematoma from lumbar puncture procedure with a history of thienopyridine use is not statistically significant (OR [95%CI], 1.904 [0.885–4.095]). Furthermore, a recent retrospective study investigating aspirin/clopidogrel use and spinal hematoma formation supports our findings [29]. Overall, this study supports the consensus that no statistically significant risk exists for epidural hematoma following lumbar puncture in patients with a history of antiplatelet use.
Epidural hematoma and anticoagulants
Multiple case studies have shown the development of epidural hematomas following heparin and other anticoagulation use [30, 31]. However, our study found no significant association between the prior use of anticoagulation therapy and risk for epidural hematoma following lumbar puncture (OR [95%CI], 1.000 [0.561–1.784]). Other studies, such as a 10-year retrospective review of central neuraxial blockade complications conducted by Moen et al. in Sweden, reported that spinal hematoma formation may be associated with anticoagulation. However, their data was inconclusive as only one-third of their patients’ outcomes were on anticoagulation [32]. Thus, no clear consensus is seen between lumbar puncture-induced spinal hematoma and anticoagulation, potentially supporting the results of our study, which showed no significant association between the risk and the outcome.
The consensus throughout current literature denotes the feasibility of performing lumbar punctures even with patients who have a prior history of using antiplatelets or anticoagulants. As discussed, the rates of complications, including but not limited to hematomas, are nominal for them to have major clinical relevance unless patients have severe pre-procedural comorbidities that need to be accounted for.
We acknowledge that there are limitations to this study. The major limitation of our study is that it is a retrospective cohort study. Furthermore, due to the limiting nature of the TriNetX database with de-identified patient information, we were not able to get specific patient-based anticoagulant and antiplatelet drug information. In addition, there is some degree of misidentification and misinformation, given the fact that TriNetX is a multi-institutional database. Nonetheless, the population-level adverse outcomes analysis of this retrospective cohort study sheds light on future research where cohorts can be designed with specific pertinent patient information to validate our current findings and further analyze the adverse outcomes of lumbar puncture procedures.
Conclusion
In summary, while most of our findings are concurrent with the current literature, no all-encompassing explanation exists that covers the entirety of the subject of risk of developing subdural hematoma and epidural hematoma following lumbar puncture for patients who have used anticoagulants or antiplatelets. Similar validation cohort studies will need to be performed to elucidate any correlations between the risk factors and potential adverse outcomes when performing lumbar punctures. This study set the stage for further investigation of the topic with the goal of bettering patient outcomes.
Acknowledgements
This declaration is not applicable.
Author contributions
JH conducted the literature review, drafted the manuscript, critically edited the manuscript, and managed the project. DRH acquired the dataset, analyzed the dataset, and critically edited the manuscript. MTS conducted the literature review, drafted the manuscript, and critically edited the manuscript. MB conducted the literature review, drafted the manuscript, and critically edited the manuscript. KK conducted the literature review, drafted the manuscript, and critically edited the manuscript. ER supervised the project and served as chief editor of the manuscript. All authors have read and approved the final manuscript.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data availability
The data is available upon request.
Declarations
Consent for publication
Authorship requirements have been met, and all authors have approved the final manuscript.
Competing interests
The authors declare no competing interests.
Consent to Participate
This declaration is not applicable.
Ethics Approval
We confirm that this study adhered to the ethical guidelines of this journal. This study includes data from the TriNetX database. The information in this database is updated daily and includes information regarding patient demographics, diagnoses, medications, and outcomes. This study utilized the TriNetX United States (U.S.) Collaborative Network, which contains information on 86,401,400 patients from 69 academic institutions. The database is federated. Thus, an Institutional Review Board approval for this study has been waived.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Doherty CM, Forbes RB (2014) Diagnostic Lumbar Puncture Ulster Med J 83(2):93–102 [PMC free article] [PubMed] [Google Scholar]
- 2.Jane LA, Wray AA (2024) StatPearls. StatPearls Publishing
- 3.Qasim A et al (2023) Unveiling the Truth: Diagnosing Bacterial Meningitis Through Repeat Lumbar Punctures - PubMed. Cureus 15(6):e40811 [DOI] [PMC free article] [PubMed]
- 4.Evans RW (1998) Complications of lumbar puncture. Neurol Clin 16(1):83–105 [DOI] [PubMed] [Google Scholar]
- 5.Park JH, Kim JY (2017) Iatrogenic Spinal Subarachnoid Hematoma after Diagnostic Lumbar Puncture. Korean J Spine 14(4):158–161 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kreppel D, Antoniadis G, Seeling W (2003) Spinal hematoma: a literature survey with meta-analysis of 613 patients. Neurosurg Rev 26(1):1–49 [DOI] [PubMed] [Google Scholar]
- 7.Choi J, Lee S, Hwangbo G (2015) Influences of spinal decompression therapy and general traction therapy on the pain, disability, and straight leg raising of patients with intervertebral disc herniation. J Phys Ther Sci 27(2):481–483 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Holl N et al (2010) [The spinal canal: from imaging anatomy to diagnosis]. J Radiol 91(9 Pt 2):950–968 [DOI] [PubMed] [Google Scholar]
- 9.Alrazooqi MK et al (2023) Traumatic Spinal Epidural Hematoma With Significant Neurologic Findings: A Case Report. Cureus 15(5):e38869 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Heestermans M et al (2022) Anticoagulants: A Short History, Their Mechanism of Action, Pharmacology, and Indications. Cells 11(20):3214 [DOI] [PMC free article] [PubMed]
- 11.Thachil J (2016) Antiplatelet therapy - a summary for the general physicians. Clin Med (Lond) 16(2):152–160 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Bougher RJ, Ramage D (1995) Spinal subdural haematoma following combined spinal-epidural anaesthesia. Anaesth Intensive Care 23(1):111–113 [DOI] [PubMed] [Google Scholar]
- 13.Raiker R et al (2021) Outcomes of COVID-19 in patients with rheumatoid arthritis: A multicenter research network study in the United States. Semin Arthritis Rheum 51(5):1057–1066 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Topaloglu U, Palchuk MB (2018) Using a Federated Network of Real-World Data to Optimize Clinical Trials Operations. JCO Clin Cancer Inf 2:1–10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Taquet M et al (2021) Cerebral venous thrombosis and portal vein thrombosis: A retrospective cohort study of 537,913 COVID-19 cases. EClinicalMedicine 39:p101061 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kim HY (2017) Statistical notes for clinical researchers: Risk difference, risk ratio, and odds ratio. Restor Dent Endod 42(1):72–76 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Andrade C (2020) Mean Difference, Standardized Mean Difference (SMD), and Their Use in Meta-Analysis: As Simple as It Gets. J Clin Psychiatry 81(5):20f13681 [DOI] [PubMed]
- 18.Horlocker TT et al (1995) Preoperative antiplatelet therapy does not increase the risk of spinal hematoma associated with regional anesthesia. Anesth Analg 80(2):303–309 [DOI] [PubMed] [Google Scholar]
- 19.Vandermeulen EP, Van Aken H, Vermylen J (1994) Anticoagulants and spinal-epidural anesthesia. Anesth Analg 79(6):1165–1177 [DOI] [PubMed] [Google Scholar]
- 20.Ruff RL, Dougherty JH (1981) Complications of lumbar puncture followed by anticoagulation. Stroke 12(6):879–881 [DOI] [PubMed] [Google Scholar]
- 21.Horlocker TT, Wedel DJ (1998) Neuraxial block and low-molecular-weight heparin: balancing perioperative analgesia and thromboprophylaxis. Reg Anesth Pain Med 23(6 Suppl 2):164–177 [DOI] [PubMed] [Google Scholar]
- 22.Domingues R et al (2016) Lumbar puncture in patients using anticoagulants and antiplatelet agents. Arq Neuropsiquiatr 74(8):679–686 [DOI] [PubMed] [Google Scholar]
- 23.Ito H et al (2024) Spinal subdural hematoma with sudden onset of paraplegia in a patient on oral anticoagulant therapy. Clin Case Rep 12(5):e8895 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Paun T et al (2022) Spinal subdural hematoma associated with lumbar puncture - A Case report. Acta Clin Croat 61(1):149–152 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kosarchuk J, Lewis C, Pham MH (2020) Massive Acute Spinal Subdural Hematoma Causing Sudden Onset Paraplegia in a Patient on Anticoagulation. Case Rep Surg 2020:8898744 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Horlocker TT, Wedel DJ, Offord KP (1990) Does preoperative antiplatelet therapy increase the risk of hemorrhagic complications associated with regional anesthesia? Anesth Analg 70(6):631–634 [DOI] [PubMed] [Google Scholar]
- 27.Maier C et al (2002) Severe bleeding following lumbar sympathetic blockade in two patients under medication with irreversible platelet aggregation inhibitors. Anesthesiology 97(3):740–743 [DOI] [PubMed] [Google Scholar]
- 28.Mayumi T, Dohi S (1983) Spinal subarachnoid hematoma after lumbar puncture in a patient receiving antiplatelet therapy. Anesth Analg 62(8):777–779 [PubMed] [Google Scholar]
- 29.Lee PW, Levy M (2019) Risk of Hematoma From Aspirin or Clopidogrel Owing to Lumbar Puncture. Mayo Clin Proc 94(8):1552–1555 [DOI] [PubMed] [Google Scholar]
- 30.Gurkanlar D et al (2007) Lumbar epidural hematoma following lumbar puncture: the role of high dose LMWH and late surgery. A case report. Neurocirugia (Astur) 18(1):52–55 [DOI] [PubMed] [Google Scholar]
- 31.Lee SJ et al (2009) Intraventricular hematoma, subarachnoid hematoma and spinal epidural hematoma caused by lumbar puncture: an unusual complication. Am J Med Sci 337(2):143–145 [DOI] [PubMed] [Google Scholar]
- 32.Moen V, Dahlgren N, Irestedt L (2004) Severe neurological complications after central neuraxial blockades in Sweden 1990–1999. Anesthesiology 101(4):950–959 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data is available upon request.