VTE prophylaxis in the neurocritically ill
Early thromboprophylaxis advised, yet neurocritical practice remains inconsistent
Full Title: Venous Thromboembolism Prophylaxis in Neurocritically Ill Patients: A Narrative Review of Timing and Agent Choice
Authors: Banerjee, O., Rodrigues, R., Adkins, L., Busl, KM.
DOI: 10.3390/jcm14134434
Article Type: Narrative Review
Clinical Question: This review examines current evidence to help clinicians determine when and how to initiate pharmacologic thromboprophylaxis (PTP) in neurocritically ill patients, balancing the prevention of venous thromboembolism (VTE) against the risk of bleeding complications.
Background & Rationale: VTE remains a significant but preventable source of morbidity in neurocritical care. Initiating PTP is a complex decision, particularly in patients with intracranial pathology or post-surgical risk. Despite the existence of older consensus statements, guidance on optimal timing and agent choice remains inconsistent, contributing to wide variability in clinical practice.
Setting: The review is based on a literature search performed in PubMed and EMBASE databases in December 2024.
Population: The review focused on patients with traumatic brain injury (TBI), intracerebral hemorrhage (ICH), subarachnoid hemorrhage (SAH), spinal or spinal cord injury (SCI), or those requiring neurosurgical intervention.
It included 90 articles, encompassing a total of 669,725 patients in relevant analyses.
Studies limited to abstracts, pre-1998 publications, and prior reviews or guidelines were excluded.
Intervention and Comparator: The review focused on the timing of PTP—typically comparing early administration (within 24 to 72 hours) versus delayed or withheld dosing. It also compared pharmacologic agents, primarily enoxaparin (ENX) versus unfractionated heparin (UFH), with occasional inclusion of other low-molecular weight heparins (LMWH) such as dalteparin.
Outcomes Assessed: Primary outcomes included VTE incidence and bleeding complications. Definitions varied, particularly for hematoma expansion (HE) in ICH and endpoint selection in SAH studies, complicating direct comparisons.
Authors’ Conclusions: Early PTP (<24–72 h) —when clinically and radiographically appropriate—appears safe and likely beneficial in reducing VTE risk across multiple neurocritical care populations. It also suggests a potential preference for LMWHs over UFH in selected patients. However, the authors highlight the predominance of retrospective data, inconsistency in outcome definitions, and ongoing practice variation as key limitations. They emphasize the need for individualised decision-making and call for higher-quality prospective studies to inform standardised practice.
⸻
Key clinical claims by authors:
Claim 1: Early PTP reduces VTE risk in neurocritically ill patients.
The review suggests that PTP initiated within 24–72 hours of injury or intervention generally decreases VTE incidence across TBI, ICH, SAH, SCI, and neurosurgical populations.
This aligns with existing guidelines recommending early PTP within 24-48 hours for ICH (especially with stable initial volumes [<25 mL] and locations) and SAH (post-securement), and 24-72 hours for SCI and neurosurgical interventions. The key is often to ensure hematoma stability via repeat CT before initiation.
Clinical application still hinges on local practice, resource availability (e.g., repeat imaging), and institutional tolerance for bleeding risk. Imaging confirmation of hemorrhage stability remains a prudent gatekeeper before initiating pharmacologic prophylaxis.
Claim 2: Early PTP is generally safe regarding bleeding complications.
Most included studies did not identify a meaningful increase in hematoma expansion or neurosurgical re-intervention with early PTP, though some exceptions—elderly glioblastoma patients, early postoperative craniotomy—deserve caution.
For neurosurgical intervention patients, early PTP (<24-72h) can be considered, but patient-specific risk vs. benefit estimation is crucial, given signals for increased repeat neurosurgery and mortality, especially within the first 72 hours post-op.
We’re unlikely to find a universal “safe time window.” The strength of this review lies in suggesting that, with appropriate imaging and clinical judgment, early prophylaxis is not inherently unsafe. But the lack of uniformity in how “bleeding” is defined or reported across studies means safety conclusions should be drawn with restraint.
Claim 3: LMWHs, particularly enoxaparin, may be preferable to UFH in certain subsets.
The data supporting a preference for one agent over another is less robust. However, there is a signal of benefit for enoxaparin in TBI, specifically in reducing VTE risk. For ICH, data is scant on agent choice. However, at least one study flagged increased ICH risk with enoxaparin in craniotomy patients.
For most patients with stable renal function and no immediate post-op concern, LMWH remains a reasonable first-line agent, with enoxaparin being a common choice. Specific guidance on dose (e.g., enoxaparin 30…40 mg q12h with CrCL ≥ 30 mL/min for TBI) is mentioned in some guidelines (WTA 2020, ACS 2024).
Claim 4: Interruption of PTP increases VTE risk.
Several retrospective studies suggest a strong link between missed prophylaxis doses and VTE incidence. In SAH patients with EVD, one study cited a 4.8-fold increased risk when even a single dose was held.
While this figure may not be broadly generalisable, the principle is sound: once PTP is deemed safe to start, consistency matters. Holding doses should be the exception, not routine—especially when the indication is well-established and bleeding risk has stabilised.
Claim 5: Patient-specific decision-making is crucial due to data gaps and conflicting results.
The review explicitly highlights data gaps and conflicting results. The interpretation of data is complicated by heterogeneity in the definition of early PTP, inconsistency in the reporting of important baseline characteristics, the definition of hematoma expansion, and the description of a standard approach to repeat imaging.
⸻
This is a narrative review, not a systematic review or meta-analysis, which inherently limits its methodological rigor compared to studies designed for quantitative synthesis. It does not employ rigorous systematic review methodologies such as explicit risk of bias assessment for each included study beyond a general categorization of study types. Instead, it aims to provide a practice-focused overview for clinicians and an updated perspective on a dynamic clinical area.
From an initial pool of over 2,300 records, 90 studies were included, including 9 RCTs, 8 prospective observational studies, and a predominance of 73 retrospective observational studies. There’s wide heterogeneity in patient selection, outcome definitions, and timing of intervention. No formal risk-of-bias assessment was reported.
Key limitations:
Study populations: wide variations in injury severity (e.g., blunt vs. penetrating TBI, small vs. large ICH volumes, elective vs. urgent spinal surgery), comorbidities, and baseline demographics. This greatly limits generalizability.
Study methodologies: significant variation in statistical methods, particularly in the chosen variables for multivariate analyses, and a large number of single-center designs.
Heterogeneity in the definition of early PTP, inconsistency in the reporting of important baseline characteristics (e.g., ICH score and GCS reported in only three and six studies, respectively), the definition of hematoma expansion (some defined by any growth and others by different levels of expansion), and the description of a standard approach to repeat imaging.
Outcome definitions and reporting: hematoma expansion, VTE events, functional outcomes and even mortality are variably defined or not reported in sufficient detail.
Safety event underreporting: bleeding complications, particularly in SCI cohorts, are relatively rare in the data—not necessarily because they don’t occur, but possibly because the studies are underpowered to detect them.
Confounding and bias: nany studies lack sufficient adjustment for confounders. Even among those using multivariate regression or propensity scores, the variables included are inconsistent.
WLST and immortal time bias: if patients with devastating injuries die or have care withdrawn before receiving PTP, it can artificially deflate complication rates in the delayed or untreated groups. Similarly, immortal time bias—where patients must survive long enough to be included in the “early PTP” arm—can inflate perceived benefit.
Significant practice changes over time affect the generalizability of older studies, especially given advancements in neurocritical care (e.g., imaging, ICP management, general ICU practices).
Taken together, the certainty of evidence supporting most of the review’s conclusions should be graded as low to very low. That doesn’t invalidate the conclusions, but it does mean they should be treated as suggestive, not prescriptive.
⸻
Implementing the review's suggested approaches for PTP in neurocritically ill patients involves several practical considerations. The current evidence suggests that:
Timing matters, but the safe window varies by diagnosis and hemorrhage dynamics. Imaging reassessment is key. Early PTP (generally within 24-72 hours) appears to be safe and effective for VTE prevention across various neurocritical illnesses, often without a significant increase in major bleeding complications.
The practice of "early repeat CT to inform PTP" implies readily available CT scanning capabilities and rapid interpretation, which may be a resource constraint in some settings.
Overcoming the "feared risk of hematoma expansion" requires clinicians to trust the data and make rapid decisions, which may necessitate educational initiatives or decision-support tools.
This is in alignment with our practice of “aggressive early prophylaxis”—initiation within 24–48 hours if repeat imaging confirms stability, although we commonly use enoxaparin 40mg once daily.
LMWH preference over UFH
Dosing of enoxaparin needs adjustment in patients with CrCl <30 mL/min, requiring careful renal function monitoring.
Consistency matters. Missed PTP doses or PTP interruption are consistently associated with an increased VTE risk across TBI, SCI, and neurosurgical SAH populations. This emphasizes the importance of consistent prophylaxis once initiated.
Decisions to hold PTP (e.g., for procedures like EVD removal) are common but lack strong guiding data, requiring careful case-by-case assessment.
A couple of extra points:
Mechanical prophylaxis remains under-discussed.
While the review focuses on pharmacologic prophylaxis, the we feel it is important to emphasize that mechanical prophylaxis (e.g., intermittent pneumatic compression - IPC) should be initiated early, especially when chemical prophylaxis is delayed or contraindicated due to high bleeding risk. This provides a crucial interim strategy.
Impact of baseline coagulopathy.
A non-insignificant percentage of neurocritically ill patients present with hemostatic failure, which drastically worsens prognosis. This underscores the importance of rapidly correcting pre-existing coagulopathy before or concurrently with initiating VTE prophylaxis, as the risk of bleeding may be higher in this subgroup, regardless of timing.
Interestingly, a post-hoc analysis of the INTERACT2 study, which was a large prospective cohort study of 754 propensity score-matched patients, found that PTP initiation within 7 days after ICH lead to…
Increased risk of major disability (mRS 3–5 at 90 days). This association was particularly observed in a subgroup of patients with NIHSS < 15.
Decreased risk of death in the overall population (though not in survivors at 48h).
The authors noted that PTP was not associated with greater rebleeding or hematoma expansion.
⸻
Our position
For clinicians managing neurocritically ill patients, the evidence largely supports initiating pharmacologic pharmacologic VTE prophylaxis relatively early (within 24-72) as a generally safe and effective strategy. This includes patients with ICH, TBI, SAH, SCI, or those recovering from neurosurgical intervention.
Crucially, personalized decision-making remains paramount. While a general "early is better" principle emerges, clinicians must carefully consider individual patient factors such as the stability of intracranial hemorrhage (often guided by repeat imaging), the specific type of neurocritical illness, and the presence of other bleeding risk factors or baseline coagulopathies. When PTP is initiated, consistency is key; holding doses increases VTE risk.
That said, the evidence base supporting this practice isn’t as robust as we’d like. The review pools mostly retrospective studies, many single-center, with variable definitions and significant heterogeneity across patient groups, timing cutoffs, and outcome measures. So while the overall signal favors early prophylaxis, we need to be mindful that biases like immortal time bias and withdrawal-of-care confounding may overstate benefit. The review acknowledges these issues, but we as clinicians need to internalize what that means at the bedside: apply these findings with caution, not dogma.
On the question of which agent, LMWH (typically enoxaparin) seems preferred in many situations—especially in TBI—but the comparative data against UFH are limited and not always consistent.
Finally, the importance of repeat imaging before starting PTP—especially after hemorrhage—is a consistent theme. And mechanical prophylaxis (e.g. IPC) still plays a key role, particularly as a bridge when pharmacologic options must be delayed.