Introduction

Isolated acute subdural haematoma (aSDH) is a common disorder, typically resulting from trauma, and has increased in prevalence in recent decades.1 aSDHs are found in up to one-third of patients with severe traumatic brain injury and are associated with an unfavourable outcome in many cases. The risk of developing SDHs is increased with predisposing factors that result in shrinkage of volume of the brain with subsequent tension on bridging veins. As such elderly patients are at the highest risk for developing chronic SDH (cSDH) with or without minor head injuries, due to the shrinking of the brain that occurs with advancing age.2 The mean age of diagnosis of cSDH is approximately 75 years,2 and the incidence ranges between 8 and 39 per 100 000 person-years.3–5 Other risk factors include alcohol use, bleeding diathesis, haemodialysis and preinjury use of anticoagulant medications. Recent data suggest that the 1-year mortality after cSDH is 30%.6–8 The combination of an ageing population, the higher incidence of disease in progressively older patients, and the associated morbidity and mortality makes cSDH an increasing concern in Canada, contributing to a significant burden on health systems.

Surgical treatment is commonly needed for large symptomatic aSDH and cSDH. aSDH causing significant mass effect and raised intracranial pressure is an indication for urgent surgery. In most patients with cSDH, the slower accumulation of blood results in brain compression several weeks after any initial injury with delayed presentation of focal neurological deficits or generalised symptoms such as headache, confusion and drowsiness.2 Surgical evacuation is the only proven treatment for patients with symptomatic cSDH. Given the liquefied nature of the blood within the collection, evacuation is typically performed by neurosurgeons using a small burr hole and subdural drain insertion.9 10 As the population continues to age, surgical evacuation of cSDH is projected to become the most common surgery of the brain.5 SDH commonly recurs after surgical drainage; even with successful evacuation of the initial haematoma, patients with cSDH are at risk for delayed recurrence. Approximately 10%–25% of patients will develop a recurrence necessitating repeat surgical evacuation within 6 months.2 11 Recurrence of cSDH has a considerable impact on patient functional outcomes and is a major determinant of poor quality of life.2 12 Therefore, minimising the risk of recurrence after surgical evacuation of cSDH is critical. Risk factors for recurrence of cSDH include bilateral cSDH,13 thick subdural membranes,11 poor intraoperative brain re-expansion,11 the presence of intracranial air on postoperative imaging13 and coagulopathy or use of anticoagulation after surgery.14–16

Atrial fibrillation and anticoagulation use are common in patients with SDH. Atrial fibrillation is the most common sustained cardiac arrhythmia in the elderly and its prevalence is expected to double by the year 2030.17 Patients with atrial fibrillation are at increased risk for cardioembolic stroke, therefore, anticoagulation is indicated for almost all patients according to the most recent Canadian Cardiovascular Society ‘CHADS-65’ Algorithm.18 19 Atrial fibrillation is one of the most common medical comorbidities in patients with cSDH, and preinjury rates of anticoagulation use in this population range between 40% and 50%.20 Direct oral anticoagulants (DOACs) are associated with a significantly reduced risk of SDHs versus vitamin K antagonists.21 Due to their improved safety profiles, DOACs are now the most prescribed anticoagulants with their usage continuing to increase.22 23 For most patients with cSDH, anticoagulation is typically held prior to drainage to minimise the risk of intraoperative and early postoperative bleeding. After surgery, the risk of recurrence of SDH must be balanced against the risk of thromboembolic events when deciding the timing of resuming anticoagulation.

There are currently no guidelines suggesting optimal timing of resumption of anticoagulation after any SDH both aSDH and cSDH and decisions about when to resume anticoagulation in patients with atrial fibrillation after aSDH or cSDH (with or without surgical evacuation/drainage) are made by clinicians on an individual patient basis without any high-quality evidence to guide this decision. Restarting anticoagulation after cSDH evacuation is associated with a higher risk of recurrent cSDH.14 24 On the other hand, it has been clearly shown that prolonged interruptions in anticoagulation for atrial fibrillation carry a significant risk of thromboembolic events secondary to loss of effective protection against thromboembolism.25–28 Two broad approaches are commonly adopted: (1) early resumption of anticoagulation (ie, approximately 30 days after initial diagnosis with or without surgery)29–31 to protect against thromboembolic events in atrial fibrillation and (2) delayed resumption of anticoagulation (ie, waiting until 90 days)24–26 to protect against early SDH reaccumulation.

We conducted a survey of practice patterns among Canadian neurosurgeons and stroke neurologists who commonly direct the resumption of postoperative anticoagulation. (2) Both neurosurgeons and stroke neurologists believed early resumption of anticoagulation was protective against the risk of perioperative stroke after drainage of cSDH. Over 95% of surveyed physicians believed that thromboembolic events contributed to the morbidity and mortality of patients with cSDH. There was substantial practice variability in the timing of resuming anticoagulation ranging from less than 1 week (25%) to beyond 3 months (5%). Importantly, most (87%) were willing to enrol their patients in a randomised controlled trial (RCT) comparing early (within 1 month) versus delayed (after 3 months) resumption of anticoagulation after SDH drainage. These findings are consistent with previous surveys on intracerebral haemorrhage1 and demonstrate clinical equipoise, supporting the need for an RCT.

Methods and analysisPatient and public involvement

We had directly involved a patient with prior SDH to read and provide active input in the concept and design of the trial prior to finalising.

Objective

The overarching objectives of this pilot RCT are to evaluate the feasibility of implementing this protocol and, specifically, to (1) establish the ability to meet prespecified criteria in recruitment, allocation, adherence to protocol and outcome assessment; (2) confirm the feasibility of implementing the study in multiple centres; (3) optimise inclusion/exclusion criteria for recruitment, retention, adherence and refinement of outcome definition; (4) estimate baseline rates of the safety outcomes for a larger trial and (5) inform the time and resources (both person-power and budgetary) that need to be considered for planning of the large-scale study. Secondary objectives of the pilot RCT are to obtain estimates of functional, safety and effectiveness outcomes to assist with future planning of the required sample size.

Study design and setting

This is a pragmatic pilot open-label RCT to evaluate the feasibility of performing a larger definitive trial. This future definitive open-label pragmatic RCT will compare two different approaches to resuming standard-dose prophylactic anticoagulation with DOACs for eligible patients after presentation with aSDH or cSDH: starting anticoagulation at 30 days vs starting at 90 days after the diagnosis of SDH. The choice of DOAC will be determined by the treating physician.

Trial status

The trial opened on 24 November 2022. At the time of manuscript submission, the status of the trial is on hold. The plan is to have a total of seven active recruitment sites across Canada. The trial will close by 30 June 2027.

Eligibility criteria

The study will enrol all adult patients over 18 years of age presenting with any newly diagnosed SDH, defined as an either acute or chronic haematoma in the subdural space diagnosed on a CT scan and recently (prior to presentation) receiving therapeutic anticoagulation (DOAC or warfarin) for stroke prophylaxis for atrial fibrillation. Patients can have surgical drainage (either burr hole or craniotomy) of the SDH prior to enrolment. The exclusion criteria are any acute SDH requiring decompressive craniectomy, mechanical heart valve or moderate to severe mitral stenosis, known chronic coagulopathy (elevated international normalized ratio (INR) >1.5 or partial thromboplastin time (PTT) >40 s despite anticoagulant reversal, thrombocytopaenia with platelet count <50×109/L) that is not amenable to reversal, more than 37 days has elapsed since initial diagnosis without recruitment into the trial, active gastroduodenal ulcer with bleeding, urogenital or respiratory tract haemorrhage and known pregnancy or breast feeding. We will also exclude patients who are known to have been previously non-compliant with anticoagulant therapy or those with a prerandomisation brain CT that reveals a recurrence of SDH requiring repeat surgical drainage.

Interventions

Eligible patients will be randomised to early (30+7 days) vs delayed (90±14 days) reinitiation of DOAC.

Group 1: early resumption of anticoagulation therapy

Patients randomised to the early intervention group will receive the standard of care DOACs at the standard dose starting at day 30+7 after diagnosis of aSDH or cSDH. The choice of DOAC is at the discretion of the treating physician.

Group 2: later resumption of anticoagulation therapy

Those patients randomised to the later intervention group will each receive the standard of care DOACs at the standard dose starting at day 90±14 after diagnosis of aSDH or cSDH. The choice of DOAC is at the discretion of the treating physician.

Outcomes and study duration

The coprimary outcome of the study is (1) recruitment rate defined as the number of patients enrolled within 1 year at each participating institution and (2) implementation of the study protocol defined as the proportion of enrolled patients who have completed follow-up assessments of functional outcome at 3 months. Functional outcomes will be estimated using the modified Rankin Scale (mRS), which will be the primary planned outcome measure in the future planned large RCT.

The secondary outcomes of the study will include the following (measured at 3 and 6 months):

Functional outcomes are defined as patient and physician-reported measures of disability, as measured by the mRS and the National Institutes of Health Stroke Scale (NIHSS). These measures are both widely applied for evaluating stroke patient outcomes (sensitive to both embolic and haemorrhagic events) and as an endpoint in randomised clinical trials for stroke.27–30 They have also been used as the primary endpoint for cSDH randomised clinical trials.31

Safety outcomes, including:

The composite outcome of clinically important intracranial haemorrhage progression after randomisation is defined by:

Any increase in volume of blood in the brain CT scan of at least 33% (similar to previously published randomised trials).32 33

Need for surgical evacuation of SDH if managed non-operatively initially or repeat surgical evacuation of SDH if managed operatively initially.

Death is directly related to enlarging SDH or new parenchymal bleeding.

Clinically important intracranial haemorrhage progression may be asymptomatic (ie, detected on scheduled CT head) or symptomatic (detected following unscheduled CT head). Changes in brain CT will be assessed using volumetric measurements performed by a centralised neuroradiologist blinded to treatment allocation.

Symptomatic bleeding in a critical non-intracranial region is defined by intraspinal, intraocular, retroperitoneal, intra-articular, pericardial, intramuscular with compartment syndrome or any other bleeding requiring transfusion of 2 units packed red blood cells.

Effectiveness outcomes

Composite outcome of the clinically important thromboembolic event after randomisation defined as any of the:

Symptomatic transient ischaemic attack: It refers to documented episodes of temporary and focal cerebral (including retinal) dysfunction of vascular origin, rapid in onset lasting up to a day and leaving no permanent clinical neurologic deficit.34 A neurological assessment and documentation of the event will be required.

Symptomatic, objectively confirmed ischaemic stroke: A documented episode consisting of acute clinical features of focal (or global) disturbance of cerebral function, lasting more than 24 hours or leading to death, with no apparent cause other than of vascular origin.35 A neurological assessment and documentation of the event will be required.

Symptomatic, objectively confirmed non-central nervous system systemic embolism will be diagnosed based on objective evidence of limb or organ ischemia most likely of cardioembolic aetiology. A clinical assessment and documentation of the event will be required.

Objectively proven venous thromboembolism: Clinically suspected acute deep vein thrombosis or pulmonary embolus diagnosed using compression ultrasonography and/or CT-pulmonary angiography. A clinical assessment and documentation of the event will be required.

Any death directly or indirectly caused by a thrombotic event (stroke or systemic arterial thromboembolism). Clinical documentation of the suspected cause of death will be required.

All safety and effectiveness outcomes will be confirmed by the adjudication committee cochaired by a thromboembolism expert and a stroke neurology expert. A schematic of the schedule of events is listed in table 1.

Table 1

Schedule of events

Consent

Before any study activities, written consent must be obtained using the current research ethics board (REB)-approved consent form. The investigator must ensure that all information given to participants or decision-makers has been approved by the local REB. Participants or decision-makers will have enough time to ask questions and get answers before deciding to participate. The investigator and study staff should not pressure anyone to join the study.

The consent form will be signed and dated by the participant or decision-maker and the person obtaining consent. The consent process will be documented in the clinical or research record. The original consent form will be kept by the site, and a copy will be given to the participant or decision-maker. Participants’ rights and welfare will be protected, and their clinical care will not be affected if they choose not to participate. Consent is an ongoing process and can be withdrawn at any time during the study.

Confidentiality

Information about study participants will be kept confidential according to the Personal Health Information Protection Act of 2004 (PHIPA) and the REB. PHIPA outlines the rules for collecting, using and disclosing PHI. Participants must consent to the collection, use and access of PHI unless the REB waives this requirement. Participants must be informed about:

What PHI will be collected?

Who will have access to it and why?

Who will use or disclose it?

Their right to revoke authorisation for the use of their PHI.

If a participant revokes authorisation, the investigator may use information collected before the revocation. Attempts should be made to obtain permission to collect at least vital status (ie, whether the participant is alive) at the end of their study period.

The investigator and research team members will have access to participant medical records and will collect only the information needed for the study. Sponsor-delegated monitors, institutional committee representatives and regulatory authority representatives will also have access to examine records for quality assurance, audits and evaluation of study safety and progress.

Recruitment

Research coordinators in each participating centre will screen patients twice weekly for eligibility criteria. Neurological and neurosurgical patients who are seen in (a) hospital in acute setting or (b) clinic in more subacute/chronic setting (by inquiring from participating site-investigators office) or (c) have undergone surgical evacuation of aSDH or cSDH. Screening will be performed by liaising with neurology or neurosurgery team and flagging patients on outpatient follow-up visits. Informed consent will be obtained from eligible patients or substitute decision-makers (SDM) if necessary due to the severity of the underlying brain dysfunction, up to day 30 after diagnosis (ie, window of enrolment). However, most potentially eligible patients will be approached for consideration of the study during their acute hospital admission and then reassessed at their day 30 follow-up visit for consent. Reason for non-enrolment of eligible patients will be recorded at each centre. Based on current population data, we expect that about 2–3 patients per month can be recruited in each centre.

Sample size

We plan to enrol 120 patients to estimate an adherence rate of 90% with a margin of error of ±5%. A sample of 120 will produce a 95% CI with bounds of 83%–95% around the hypothesised target of 90%. Based on local experience with a similar patient population (NCT03559114), we assume that about 70% of eligible patients will consent to participation in the trial. We, therefore, expect that 170 eligible patients will need to be screened to achieve a randomised sample of 120 patients, representing about 10% of the final sample size that we anticipate requiring for the full RCT in keeping with recommendations for pilot trials.36 We designed this pilot study with a conservative assumption that at least 12 patients can be recruited per centre per year. This number should be feasible to accrue within a 2.5-year funding period (1.5 years of active recruitment considering some staggered activation of centres) based on the expected number of recruitments at seven sites. All seven proposed pilot trial sites are academic tertiary care centres with about 50–70 cSDH admissions per year of which 40% are typically on anticoagulation.4 5

Randomisation and blinding

All eligible patients receive a brain CT scan between 2 and 4 weeks after initial diagnosis as part of the standard of care. If there is no safety concern, eligible patients will be randomised at 30+7 days of initial diagnosis of aSDH or cSDH to one of the two groups. Randomisation will be performed via web-based randomisation system in a 1:1 ratio and using variable (4–6) block sizes to avoid substantial imbalances in number of patients assigned to each group. This will be further stratified by study site (to ensure similar allocation to study groups within each centre), and aSDH versus cSDH.

All patients, physicians and study investigators are aware of the randomised treatment allocation, but the outcomes assessors and adjudication committee will remain blinded to treatment allocation. We will use intention-to-treat principle in analysing the outcomes. Enrolment bias is minimised by maintaining allocation concealment using block randomisation.

In case unblinding is necessary, investigators will be referred to our approach to unblinding (online supplemental appendix 1).

Data analysis planData collection and management

The baseline assessment is conducted 30+7 days after diagnosis of aSDH or cSDH. Routine baseline clinical information will be recorded including key comorbidities, anticoagulation use, mRS and NIHSS. In addition to the visits listed below, participants will be contacted by the research team by phone every month to enquire about safety and effectiveness outcomes. The results of the clinical CT, clinical outcomes, mRS and NIHSS are recorded at 90±14 days and 180±28 days.

If, for any reason, a participant discontinues their allocated treatment, every effort will be made to continue to collect the above outcomes measures. Participants or SDMs can also request to permanently withdraw from the study. However, every effort should be made to obtain permission to document the reason for withdrawal and to collect participant outcomes up to the protocol-described end of participant follow-up period, where possible. Any data collected prior to the withdrawal of consent will be retained. All source records and source data will be maintained by the site investigator and preserved as stipulated by the regulatory authorities.

Statistical methods

Simple descriptive statistics will be used to report the rate of recruitment, secondary outcomes, baseline demographics and clinical variables. Frequencies, proportions and their 95% CI will be used to report the implementation of study variables. For categorical variables, we use frequencies, proportions and their 95% CI.

Monitoring and quality assurance

Data for the study will be centrally stored and managed by the Centre for Clinical Trial Support as the coordinating centre. To ensure the quality of study data, quality assurance and control systems will be implemented using validated electronic quality control checks within the electronic data capture system. These verification measures will identify missing data, inconsistencies and/or data anomalies. Both electronic and manual queries will be generated for resolution and review by sites. Access to secure and validated electronic systems used for the purposes of this study will be controlled by the sponsor.

In order to mitigate the risks, the protocol incorporates strict safety criteria, four preplanned interim Data and Safety Monitoring Board (DSMB) reviews for safety and incorporates measures to screen for both cSDH recurrence and thromboembolic events during the intervention period. Please refer to online supplemental appendix 2 for a detailed DSMB charter.

Study audit

Site monitoring ensures the safety and rights of study participants and verifies the accuracy and completeness of study data. The sponsor is responsible for monitoring, which is based on the approved protocol, regulations and institutional policies. The monitoring plan outlines the frequency, procedures and level of site monitoring activities. Monitoring is performed remotely, and findings are reported to the site study team and investigator. The institution and/or local REB may conduct independent audits. The investigator must provide access to all study-related documents and facilities for monitoring, audits and inspections.

Concamitant medications

Contraindicated medications are listed in the exclusion criteria. Relevant concomitant medications and treatments will be recorded at baseline, 90 and 180 days. For patients who were on concomitant antiplatelet therapy prior to surgery, the antiplatelet therapy can be resumed at the discretion of the treating physician and will be recorded. While the available evidence suggests that resumption of antiplatelet agents is not associated with increased risk of cSDH, there are very few indications for patients to be on both an anticoagulant and an antiplatelet agent.