CLOTS 3
Effectiveness of intermittent pneumatic compression in reduction of risk of deep vein thrombosis in patients who have had a stroke (CLOTS 3): a multicentre randomised controlled trial
Bottom Line
In immobile stroke patients, thigh-length intermittent pneumatic compression applied for up to 30 days significantly reduced proximal DVT (8.5% vs 12.1%; adjusted OR 0.65, p=0.001) with a trend toward improved 6-month survival and only a small excess of skin breaks.
Major Points
- Largest RCT of IPC in stroke patients (N=2876) with a clear reduction in proximal DVT within 30 days (absolute risk reduction 3.6%, NNT ≈28).
- Benefit was consistent across prespecified subgroups including ischemic vs haemorrhagic stroke, stroke severity, delay from onset, and background heparin/thrombolysis use.
- Improvement in 6-month survival (22.3% vs 25.1%, OR 0.85, p=0.059) is biologically plausible and did not occur at the cost of major adverse effects (unlike heparin/LMWH which increase bleeding).
- IPC caused more skin breaks (3.1% vs 1.4%, p=0.002) but few were directly attributed to the device; no excess in falls, fractures, or lower-limb ischemia.
- Established IPC as a guideline-recommended standard-of-care for VTE prophylaxis in immobile stroke patients (adopted by AHA/ASA and NICE).
Design
Study Type: Multicentre, parallel-group, randomised controlled trial
Randomization: 1
Blinding: Open-label treatment with masked outcome assessment (CDU technicians blinded to allocation; caregivers and patients not masked)
Enrollment Period: December 8, 2008 – September 6, 2012
Follow-up Duration: 6 months
Centers: 94
Countries: United Kingdom
Sample Size: 2876
Analysis: Intention-to-treat, with adjustment for minimisation variables (logistic regression)
Inclusion Criteria
- Admission to hospital within 3 days of acute stroke
- Immobile at enrollment (unable to walk to the toilet without the help of another person)
- Written informed consent from patient or proxy
- Enrollment day 0 through day 3 of admission
Exclusion Criteria
- Age younger than 16 years
- Subarachnoid haemorrhage
- Contraindication to IPC: active dermatitis
- Contraindication to IPC: leg ulcers
- Contraindication to IPC: severe leg oedema
- Severe peripheral vascular disease
- Congestive cardiac failure
Arms
| Field | IPC | Control |
|---|---|---|
| Intervention | Thigh-length Kendall SCD Express sequential compression sleeves (Covidien) applied to both legs, worn continuously day and night for up to 30 days or until independently mobile, discharged, second CDU completed, or adverse events required removal | Usual care without IPC; background thromboprophylaxis (heparin/LMWH, graduated stockings) at clinician discretion |
| Duration | Up to 30 days (median 9 days [IQR 3–22], mean 12.5 days; perfect adherence in 31%) | Follow-up 30 days plus 6-month outcomes |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Symptomatic or asymptomatic DVT in the popliteal or femoral (proximal) veins on screening compression duplex ultrasound, or any imaging-confirmed symptomatic proximal DVT, within 30 days of randomisation (primary analysis excludes patients who died before any primary outcome and those without a screening CDU) | Primary | 12.1% (174/1438); adjusted 14.0% (174/1245) | 8.5% (122/1438); adjusted 9.6% (122/1267) | 0.65 | 0.001 |
| Death by 30 days | Secondary | 13.1% (189/1438) | 10.8% (156/1438) | 0.8 | 0.057 |
| Symptomatic proximal DVT (30 days) | Secondary | 3.4% (49/1438) | 2.7% (39/1438) | 0.79 | 0.269 |
| Asymptomatic proximal DVT (30 days) | Secondary | 8.7% (125/1438) | 5.8% (83/1438) | 0.65 | 0.003 |
| Symptomatic DVT proximal or calf (30 days) | Secondary | 6.3% (90/1438) | 4.6% (66/1438) | 0.72 | 0.045 |
| Any DVT (30 days) | Secondary | 21.1% (304/1438) | 16.2% (233/1438) | 0.72 | 0.001 |
| Confirmed pulmonary embolism (30 days) | Secondary | 2.4% (35/1438) | 2.0% (29/1438) | 0.83 | 0.453 |
| Any DVT or death (30 days) | Secondary | 32.8% (472/1438) | 26.2% (377/1438) | 0.72 | <0.0001 |
| Any DVT, PE, or death (30 days) | Secondary | 34.1% (491/1438) | 27.2% (391/1438) | 0.72 | <0.0001 |
| Dead by 6 months | Secondary | 25.1% (361/1438) | 22.3% (320/1438) | 0.85 | 0.059 |
| Any DVT at 6 months | Secondary | 21.7% (312/1438) | 16.7% (240/1438) | 0.72 | 0.001 |
| Any symptomatic DVT at 6 months | Secondary | 7.0% (101/1438) | 5.4% (77/1438) | 0.75 | 0.061 |
| Any confirmed PE at 6 months | Secondary | 3.4% (49/1438) | 2.9% (42/1438) | 0.86 | 0.463 |
| Any death, DVT, or PE at 6 months | Secondary | 43.5% (626/1438) | 36.6% (526/1438) | 0.74 | <0.0001 |
| Skin breaks (any, 30 days) | Adverse | IPC 3.1% (44/1438) vs no IPC 1.4% (20/1438); OR 2.23 (1.31–3.81), p=0.002 | |||
| Skin breaks attributed to IPC | Adverse | IPC 0.7% (10/1438) vs no IPC 0% | |||
| Lower limb ischaemia or amputation | Adverse | IPC 0% vs no IPC 0.1% (2/1438) | |||
| Falls with injury (30 days) | Adverse | IPC 2.3% (33/1438) vs no IPC 1.7% (24/1438); OR 1.39 (0.82–2.37), p=0.221 | |||
| Falls with injury attributed to IPC | Adverse | 1 patient (0.1%) | |||
| Fractures within 30 days | Adverse | IPC 0.3% (4/1438) vs no IPC 0.3% (4/1438) | |||
Subgroup Analysis
No significant interactions in prespecified subgroups: delay from stroke onset, leg weakness, stroke severity (validated prognostic model), DVT risk category, use of heparin/warfarin/thrombolysis at enrollment, stroke type (haemorrhagic vs ischaemic), or IPC sleeve type (Original vs Comfort). Effect was consistent, including a similar benefit in patients with haemorrhagic stroke.
Criticisms
- Open-label design: caregivers and patients were not masked, which could bias use of background thromboprophylaxis and detection/reporting of secondary outcomes and adverse events.
- Imperfect masking of ultrasound technicians: 156 (5%) patients attended the CDU wearing IPC sleeves, potentially unmasking the assessor.
- Moderate adherence: mean adherence was 59% and only 31% of patients achieved 'perfect' adherence to prescribed IPC use.
- Calf veins were not fully visualised in ~47% of screening ultrasounds, so isolated calf DVTs may have been missed.
- Pulmonary embolism was not systematically screened for; autopsy rates were very low, limiting attribution of cause of death.
- Trial powered for DVT reduction, not survival; the 6-month mortality signal (OR 0.85, p=0.059) requires further study for confirmation.
- Conducted exclusively in the UK, which may limit generalisability to different healthcare systems and populations.
Funding
National Institute for Health Research (NIHR) Health Technology Assessment (HTA) programme, UK; Chief Scientist Office of the Scottish Government; Covidien (Massachusetts, USA) donated IPC devices and sleeves but had no role in study design, analysis, or reporting.
Based on: CLOTS 3 (Lancet, 2013)
Authors: CLOTS Trials Collaboration; Dennis M, Sandercock P, Reid J, ..., Murray G
Citation: Lancet 2013;382(9891):516–524. DOI: 10.1016/S0140-6736(13)61050-8
Content summarized and formatted by NeuroTrials.ai.