SWITCH
Decompressive craniectomy plus best medical treatment versus best medical treatment alone for spontaneous severe deep supratentorial intracerebral haemorrhage: a randomised controlled clinical trial
Clinical Question
Does decompressive craniectomy plus best medical treatment improve clinical outcome at 6 months for people with spontaneous severe deep intracerebral haemorrhage compared to best medical treatment alone?
Study Overview
Objective
Assess whether decompressive craniectomy plus best medical treatment improves functional outcome at 6 months compared with best medical treatment alone in patients with severe deep supratentorial intracerebral hemorrhage.
Study Summary
In patients with spontaneous severe deep supratentorial intracerebral hemorrhage, decompressive craniectomy plus best medical treatment showed a trend toward better outcomes (lower mRS 5–6) compared with best medical treatment alone, but the result did not reach statistical significance. The trial was stopped early due to lack of funding.
Intervention
Multicenter, randomized, open-label, assessor-blinded trial in 42 centers across Europe. Patients aged 18–75 years with ICH involving basal ganglia or thalamus were randomized to decompressive craniectomy (≥12 cm diameter, without hematoma evacuation) plus best medical treatment vs. best medical treatment alone. N=197; median hematoma volume ~57 mL; follow-up at 180 and 365 days.
Bottom Line
The SWITCH trial provides weak evidence that decompressive craniectomy plus best medical treatment might be superior to best medical treatment alone in people with severe deep intracerebral haemorrhage, with a trend toward improved outcome but survival remained associated with severe disability in both groups.
Major Points
- The trial was stopped early due to lack of funding, resulting in being underpowered for the primary endpoint.
- In the intention-to-treat analysis, 42 (44%) of 95 participants in the decompressive craniectomy group and 55 (58%) of 95 in the best medical treatment group had an mRS of 5–6 at 180 days (adjusted risk ratio [aRR] 0.77, 95% CI 0.59 to 1.01, adjusted risk difference −13%, p=0.057).
- Ordinal mRS analysis favored surgery (common odds ratio 0.57, 95% CI 0.34 to 0.97).
- Mortality at 180 days was lower with surgery: 17% vs 27% (aRR 0.61, 95% CI 0.36 to 1.01).
- Severe adverse events occurred in 42 (41%) of 103 receiving surgery vs 41 (44%) of 94 receiving best medical treatment alone (safety set).
- Median hospital stay was shorter in the surgical group: 19.5 vs 23.5 days (time ratio 0.74, 95% CI 0.57 to 0.95; p=0.018).
- Survival remained associated with severe disability in most cases.
Design
Study Type: Multicentre, randomised, open-label, assessor-blinded trial
Randomization: 1
Blinding: Assessor-blinded for outcome assessment; investigators and core lab staff masked to allocation and outcomes.
Enrollment Period: October 6, 2014, to April 30, 2023
Follow-up Duration: 6 months (primary outcome), up to 12 months total
Centers: 42
Countries: Austria, Belgium, Finland, France, Germany, Netherlands, Spain, Sweden, Switzerland
Sample Size: 201
Randomisation Window: Randomisation was performed within 66 h after symptom onset; decompressive craniectomy was performed no later than 6 h after randomisation.
Analysis: Intention-to-treat; Mantel-Haenszel risk ratios for binary outcomes; proportional odds regression for ordinal outcomes; linear models for continuous data; flexible parametric accelerated failure time model (time ratio) for length of hospital stay; multiple imputation for sensitivity.
Inclusion Criteria
- Adults (18–75 years)
- Spontaneous severe deep intracerebral haemorrhage (basal ganglia or thalamus, may extend to lobes, ventricles, or subarachnoid space)
- NIHSS ≥10, ICH volume ≥30 mL, and GCS <14
Exclusion Criteria
- People with a GCS score of 4–7 were deliberately excluded (per authors: aim was to reduce disability and avoid an outcome of mRS 5, rather than to reduce mortality at any cost)
- Full inclusion and exclusion criteria are listed in the trial appendix (p 6) and not reproduced in the main paper
Arms
| Field | Control | Decompressive craniectomy plus best medical treatment |
|---|---|---|
| Intervention | Best medical management per standard protocols | Decompressive craniectomy (≥12 cm diameter) without haematoma evacuation plus best medical treatment; delayed cranioplasty allowed. |
| Duration | 180 days (primary outcome) | 180 days (primary outcome) |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| mRS 5–6 at 180 days in the intention-to-treat population | Primary | 58% | 44% | 0.77 | 0.057 |
| Ordinal mRS at 180 days | Secondary | Median 5 (IQR 4–6) | Median 4 (IQR 4–5) | 0.57 | 0.039 |
| Mortality at 180 days | Secondary | 27% | 17% | 0.61 | 0.065 |
| Length of hospital stay (median days) | Secondary | 23.5 | 19.5 | 0.74 | 0.018 |
| Safety Set | Adverse | Analysed in all randomised participants according to treatment actually received: n=103 received decompressive craniectomy plus best medical treatment; n=94 received best medical treatment alone (total safety set n=197). | |||
| Any serious adverse event | Adverse | 41/94 (44%) | 42/103 (41%) | 0.77 | |
| Infections and infestations | Adverse | 11/94 (12%) | 12/103 (12%) | 1.00 | |
| Nervous system disorders | Adverse | 17/94 (18%) | 17/103 (17%) | 0.85 | |
Criticisms
- Underpowered due to early termination after only 201 of planned 300 patients enrolled.
- Assumed large treatment effect may have masked smaller benefits.
- Open-label design, with possible bias from participant awareness.
- Variability in surgical technique and evolving care standards over long enrollment period.
- Limited female representation (~33%) and lack of ethnic data.
- Data from imaging core lab on specific anatomical involvement not provided.
Subgroup Analysis
No treatment effect heterogeneity across prespecified subgroups (age, haematoma size, etc.).
Funding
Swiss National Science Foundation, Swiss Heart Foundation, Inselspital Stiftung, Boehringer Ingelheim
Based on: SWITCH (The Lancet, 2024)
Authors: Jürgen Beck, Christian Fung, Daniel Strbian, ..., Urs Fischer
Citation: Lancet 2024; 403: 2395-404
Content summarized and formatted by NeuroTrials.ai.