AUTHORS
Cecily Allen, MD, Shing Chao, MD2, Matthew Lankiewicz, MD3, Morgan Walker, MD4, Jess Chasler, PharmD5, Peter Dziedzic6, Michael Streiff, MD7, Sameer S Kadri, MD, MS8
AFFILIATIONS
1 National Institutes of Health, Bethesda, Maryland, United States
2 Johns Hopkins University, Baltimore, Maryland, United States
3 Johns Hopkins University, Baltimore, Maryland, United States
4 National Institutes of Health, Bethesda, Maryland, United States
5 Johns Hopkins University, Baltimore, Maryland, United States
6 Johns Hopkins University, Baltimore, Maryland, United States
7 Johns Hopkins University, Baltimore, Maryland, United States
8 National Institutes of Health, Bethesda, Maryland, United States
ABSTRACT
Therapeutic anticoagulation has a wide range of indications and increasing utilization. It is estimated that 2%–25% of patients undergoing neurosurgery are on anticoagulation. Therapeutic anticoagulation around neurosurgery presents a clinical challenge with balancing thrombotic risk with intracranial hemorrhage. Data regarding the optimal timing of anticoagulation initiation following neurosurgery are scarce, and guidelines for perioperative neurosurgical anticoagulation are lacking. We aimed to describe practice patterns and patient outcomes of anticoagulation initiation using large-scale retrospective data for cranial and spinal neurosurgeries.
A retrospective chart review was conducted from 2016–2024 for patients ≥ 18 years old undergoing inpatient cranial or spinal surgery at a tertiary referral academic medical center. We evaluated the number of patients who received postoperative anticoagulation, postoperative day of initiation, agent of choice and patient outcomes (bleeding, thrombosis, and reoperation). Patients were matched based on the specific surgery at a 1:2 ratio of therapeutic anticoagulation to prophylactic or no anticoagulation. Outcomes for patients were abstracted for the first 30 postoperative days. Bleeding events were defined using the International Society on Thrombosis and Haemostasis (ISTH) definitions with major bleed subcategory of intracranial hemorrhage (ICH). Indications for anticoagulation were classified as venous thromboembolism (VTE), atrial fibrillation/atrial flutter (AF), antiphospholipid syndrome (APS), heparin induced thrombocytopenia (HIT), or mechanical valve (MV). Continuous variables were summarized and compared using Wilcoxon rank-sum or Kruskal-Wallis tests; categorical variables were compared using Fisher's exact test.
A total of 1,256 patients underwent neurosurgery. Median age was 60.2 (IQR 17) and 49 % were female. There were 690 (56%) cranial and 552 (44%) spinal surgeries. Postoperative anticoagulation intensity was as follows: prophylactic dose only, 809 (64.4%); therapeutic initially (no preceding prophylactic dose), 52 (4.1%); started on prophylactic but transitioned to therapeutic, 317 (25.2%); no anticoagulation, 78 (6.2%). Among patients initiated on therapeutic anticoagulation after neurosurgery (n=369), indications (not mutually exclusive) were: acute VTE, 199 (53.9%); history of VTE, 113 (30.6%); atrial fibrillation, 95 (25.7%); HIT, 7 (1.9%); mechanical valve, 6 (1.6%); APS, 1 (0.3%). The median postoperative day of therapeutic anticoagulation initiation was 7 (IQR=10). The most common initial anticoagulant was unfractionated heparin, 187 (50.7%), followed by direct oral anticoagulants, 118 (32.0%); enoxaparin 40 (10.8%); warfarin, 14 (3.8%); and direct thrombin inhibitor, 10 (2.7%). Bleeding occurred in 36/369 (9.8%) therapeutically anticoagulated patients: 2 (5.6%) minor, 10 (27.8%) clinically relevant non-major, 15 (41.7%) major bleed (non-ICH), and 9 (25.0%) ICH. In patients who had a pre-operative indication for anticoagulation, there was no statistically significant difference in rates of major bleeding or ICH between patients who started therapeutic anticoagulation ≤7 days postoperatively (3 (3.2%)) versus >7 days postoperatively 7 (7.8%) (p=0.3). In this same group, thrombotic rates were comparable at ≤7 days postoperative 15 (16%) vs >7 days postoperatively 13 (14%). In all patients, ICH rates were higher in patients who received therapeutic anticoagulation (25/369, 6.8%) compared with those on prophylactic dosing only (21/809, 2.6%) (p=0.001). The median day of bleeding events after initiation of either prophylactic or therapeutic anticoagulation was 6 days.
This is the first large-scale study evaluating practice patterns and outcomes of anticoagulation initiation after neurosurgery. Among patients who initiated therapeutic anticoagulation, acute VTE was the most common indication, with a median time to initiation of postoperative day 7. Bleeding occurred in approximately 1 in 10 therapeutically anticoagulated patients, most commonly major bleeding or ICH. Major bleeding and ICH rates were comparable between patients who initiated therapeutic anticoagulation ≤7 versus >7 days postoperatively. These findings may help inform individualized, risk-based timing of anticoagulation initiation in the postoperative neurosurgical setting.
DISCLOSURES
Cecily Allen, MD: No relevant conflicts of interest to declare.