Guidelines for the Early Management of Patients With Acute Ischemic Stroke

This is not a single experiment but a consensus-based guideline. The writing committee reviewed all published literature on acute ischemic stroke management and graded the strength of evidence for each intervention. The key interventions evaluated include:

**Reperfusion therapies:** Intravenous tPA (alteplase), intra-arterial tPA, mechanical thrombectomy (clot retrieval)

**Supportive care:** Blood pressure management, oxygen supplementation, glucose control, temperature management

**Diagnostic procedures:** Brain imaging (CT, MRI), vascular imaging (CT angiography, MR angiography), cardiac monitoring

**Organisational systems:** Stroke unit care, telemedicine, emergency medical services protocols

The primary outcome was **reduced disability at 3 months**, measured by the modified Rankin Scale (mRS, 0–6 scale where 0 = no symptoms, 6 = death). Secondary outcomes included mortality, symptomatic intracranial haemorrhage (bleeding in the brain caused by treatment), and functional independence.

The guideline is based on data from multiple clinical trials and observational studies involving **tens of thousands of patients** across dozens of countries. Key populations from the landmark trials include:

**NINDS tPA trial (1995):** 624 patients with acute ischemic stroke, mean age 67 years, treated within 3 hours of symptom onset

**ECASS III trial (2008):** 821 patients, mean age 65 years, treated within 3–4.5 hours

**Multiple thrombectomy trials (2015 onwards):** ~1,500 patients with large vessel occlusions, mean age 68 years, treated within 6–24 hours depending on imaging criteria

The guideline applies to **adults aged 18+ with acute ischemic stroke** (not haemorrhagic stroke). Excluded populations include patients with recent major surgery, active bleeding, coagulopathy, or very mild/rapidly improving symptoms.

The guideline uses standardised clinical scales and imaging protocols:

**National Institutes of Health Stroke Scale (NIHSS):** 0–42 scale, higher = worse neurological deficit. Used to quantify stroke severity at presentation and track changes.

**Modified Rankin Scale (mRS):** 0–6 scale, the primary outcome measure for most stroke trials. Scores 0–2 indicate functional independence (able to live without assistance).

**Symptomatic intracranial haemorrhage (sICH):** Defined as any intracranial bleeding with neurological worsening (typically ≥4-point increase on NIHSS). This is the major safety outcome for tPA treatment.

**Door-to-needle time:** Minutes from hospital arrival to tPA administration. Target is ≤60 minutes.

**Door-to-puncture time:** Minutes from arrival to start of endovascular thrombectomy. Target is ≤90 minutes.

**Imaging:** Non-contrast CT to rule out haemorrhage, CT angiography to identify large vessel occlusions, MRI with diffusion-weighted imaging for early ischemic changes.

**Study design:** This is a **clinical practice guideline**, not a primary research study. The writing committee (23 members appointed by the American Stroke Association) used a formal consensus process with systematic literature review. They graded each recommendation using the American Heart Association's Level of Evidence (LOE) and Class of Recommendation (COR) system:

**Class I:** Benefit >>> risk (strong recommendation, "should be done")

**Class IIa:** Benefit >> risk (moderate recommendation, "reasonable to do")

**Class IIb:** Benefit ≥ risk (weak recommendation, "may be considered")

**Class III:** No benefit or harm (recommendation against)

**Level of Evidence A:** Multiple randomised trials

**Level of Evidence B:** Single randomised trial or multiple non-randomised studies

**Level of Evidence C:** Expert opinion, case studies, or standard of care

**What this design can prove:** Guidelines can synthesise high-quality evidence into actionable recommendations. When a recommendation is Class I, Level A (e.g., tPA within 3 hours), it reflects consistent results from multiple randomised trials with thousands of patients. These recommendations have strong internal validity and generalisability to similar populations.

**What this design cannot prove:** Guidelines are not experiments. They cannot establish causality for interventions where evidence is weak (Level B or C). Many recommendations are based on expert opinion (Level C) because randomised trials are impractical or unethical (e.g., "don't give tPA to patients with active bleeding"). Guidelines also cannot account for individual patient variability—they provide population-level averages, not personalised predictions.

**Major methodological weaknesses:**

**Conflict of interest:** Many committee members had financial ties to pharmaceutical companies (tPA manufacturers, device companies). The guideline reports these conflicts but does not quantify their impact on recommendations.

**Publication bias:** The literature review likely overrepresents positive trials. Negative trials are less likely to be published.

**Outdated evidence:** This guideline was published in 2013. Major advances since then include extended thrombectomy windows (up to 24 hours based on DAWN/DEFUSE-3 trials, 2018) and tenecteplase as an alternative to alteplase. The 2019 update and 2022 guidelines supersede this document.

**Lack of patient-centred outcomes:** The guideline focuses on mRS at 90 days, but patients may value other outcomes (quality of life, cognitive function, caregiver burden) that are not systematically addressed.

The guideline organises findings by intervention category. Below are the strongest recommendations with supporting evidence:

**Intravenous tPA (alteplase):**

**Class I, Level A:** tPA 0.9 mg/kg (max 90 mg) within 3 hours of symptom onset. Number needed to treat (NNT) = 8 for one additional patient with excellent outcome (mRS 0–1 at 90 days). Number needed to harm (NNH) = 33 for symptomatic intracranial haemorrhage.

**Class I, Level B:** tPA within 3–4.5 hours in eligible patients (ECASS III criteria: age ≤80, no diabetes + prior stroke, NIHSS ≤25, not on anticoagulants). NNT = 14 for excellent outcome.

**Absolute benefit:** In the NINDS trial, 39% of tPA-treated patients had mRS 0–1 at 90 days vs. 26% of placebo (absolute risk reduction 13%, p=0.008).

**Risk of sICH:** 6.4% with tPA vs. 0.6% with placebo in NINDS (p<0.001). Mortality at 90 days was not significantly different (17% vs. 21%, p=0.30).

**Endovascular thrombectomy:**

**Class I, Level B:** For patients with large vessel occlusion (internal carotid artery or proximal middle cerebral artery) treated within 6 hours. Based on early trials (MR CLEAN, ESCAPE, EXTEND-IA) published after the 2013 guideline but incorporated into the 2015 update.

**Effect size:** In MR CLEAN (2015), 32.6% of thrombectomy patients achieved functional independence (mRS 0–2) vs. 19.1% with medical therapy alone (adjusted odds ratio 2.16, 95% CI 1.39–3.38).

**Blood pressure management:**

**Class I, Level B:** Before tPA, blood pressure should be lowered to <185/110 mmHg. After tPA, maintain <180/105 mmHg for 24 hours.

**Class IIb, Level C:** In patients not receiving tPA, permissive hypertension (up to 220/120 mmHg) is reasonable in the first 24–48 hours unless other contraindications exist. No randomised trial data support aggressive lowering.

**Anti-thrombotic therapy:**

**Class I, Level A:** Aspirin 325 mg within 24–48 hours of stroke onset (not within 24 hours of tPA). Reduces recurrent stroke risk by ~1% absolute (NNT = 100).

**Class III, Level A:** Do not use heparin or other anticoagulants in the acute phase (within 24 hours). No benefit and increased bleeding risk.

**Supportive care:**

**Class I, Level C:** Maintain oxygen saturation ≥94% with supplemental oxygen if needed. No evidence for routine oxygen in non-hypoxic patients.

**Class IIa, Level C:** Treat hyperglycemia >180 mg/dL with insulin. Avoid hypoglycemia. Target glucose 140–180 mg/dL.

**Class IIb, Level C:** Induced hypothermia is not recommended outside clinical trials. No proven benefit in stroke.

To translate the numbers into plain English:

**tPA within 3 hours:** For every 8 patients treated, one additional patient will be disability-free at 3 months compared to placebo. That means a 13% absolute increase in excellent outcomes. However, for every 33 patients treated, one will have a symptomatic brain bleed caused by the treatment—though most of these patients survive, and overall mortality is not increased.

**tPA within 3–4.5 hours:** The benefit is smaller (NNT = 14), meaning the treatment effect is roughly half as strong in this window. The risk of bleeding remains similar.

**Thrombectomy:** For every 7–8 patients with large vessel occlusion who undergo clot retrieval, one additional patient achieves functional independence. This is a large effect—roughly doubling the chance of a good outcome compared to medical therapy alone.

**Aspirin:** The benefit is modest—1% absolute reduction in recurrent stroke. This is roughly equivalent to preventing one stroke per 100 patients treated.

The authors acknowledge several limitations, and critical readers should note additional concerns:

**Acknowledged by authors:**

Many recommendations are based on limited data (Level C evidence)

The guideline cannot address all clinical scenarios or patient comorbidities

Rapidly evolving evidence means some recommendations may become outdated

Conflict of interest disclosures are provided but not analysed for impact on recommendations

**Critical reader observations:**

**Outdated evidence:** This 2013 guideline is now superseded. Major advances include: (1) thrombectomy extended to 24 hours for selected patients (DAWN trial, 2018), (2) tenecteplase as an alternative to alteplase (EXTEND-IA TNK, 2018), (3) direct oral anticoagulants for secondary prevention, (4) COVID-19 stroke management protocols.

**Population homogeneity:** The landmark tPA trials enrolled predominantly white patients (68% in NINDS). Efficacy and safety in non-white populations, especially those with higher rates of hypertension and diabetes, may differ.

**Real-world generalisability:** Trial patients were treated at expert centres with rapid protocols. Community hospitals may have longer door-to-needle times and higher complication rates. The guideline's targets (door-to-needle ≤60 minutes) are aspirational and achieved in only ~50% of US hospitals.

**Patient selection complexity:** The guideline's eligibility criteria for tPA are extensive (12 absolute contraindications, 15 relative contraindications). In practice, only 5–10% of stroke patients receive tPA, partly due to late arrival and partly due to these restrictions.

**Industry influence:** The writing committee included members with financial ties to Genentech (tPA manufacturer), Boehringer Ingelheim, and device companies. While conflicts were disclosed, their impact on recommendation strength (especially for tPA) cannot be quantified.

**No cost-effectiveness analysis:** The guideline does not address the economic implications of recommendations. tPA costs ~$5,000 per dose, thrombectomy ~$15,000–$30,000 per procedure. Cost-effectiveness ratios vary widely by healthcare system.

For someone running their own n=1 experiment, this guideline is not directly applicable—it's designed for healthcare systems, not individuals. However, the underlying principles can inform personal health optimisation:

### What to test

**Stroke recognition and response time:** Test your ability to recognise stroke symptoms using the FAST acronym (Face drooping, Arm weakness, Speech difficulty, Time to call 911). Measure your response time in a simulated scenario.

**Blood pressure variability:** If you have hypertension, test the effect of medication timing (morning vs. evening dosing) on 24-hour blood pressure variability. Use a home BP monitor with automatic readings every 30 minutes for 24 hours.

**Glucose control:** If you have diabetes or prediabetes, test the effect of a low-glycaemic-index breakfast on postprandial glucose spikes. Use a continuous glucose monitor (CGM) for 7–14 days.

### Minimum meaningful duration

**Blood pressure variability:** 7 days per condition (morning vs. evening dosing) with at least 3 days of washout between conditions. Blood pressure adapts within 2–3 days of medication changes.

**Glucose response:** 3–5 days per dietary intervention. Postprandial glucose responses stabilise within 24–48 hours of consistent eating patterns.

**Stroke recognition training:** One-time test, but retention should be reassessed at 1 month and 6 months.

### What to measure (specific metrics)

**Blood pressure:** Mean systolic BP, diastolic BP, and standard deviation (variability) over 24 hours. Target: mean <130/80 mmHg, variability <12 mmHg standard deviation.

**Glucose:** Peak postprandial glucose (mg/dL), time to peak (minutes), area under the curve (AUC) for 2 hours post-meal. Target: peak <180 mg/dL, return to baseline within 2 hours.

**Stroke recognition:** Time to correctly identify all FAST symptoms (seconds), accuracy (proportion correct out of 10 simulated scenarios).

### Key confounds to control for

**Time of day:** Blood pressure and glucose have circadian rhythms. Measure at the same times each day.

**Medication timing:** Take all medications at the same time relative to meals and measurements.

**Physical activity:** Record exercise type, duration, and time relative to measurements. Exercise lowers BP for 4–6 hours post-activity.

**Sleep:** Poor sleep increases BP variability and impairs glucose metabolism. Track sleep duration and quality (use a sleep diary or wearable).

**Caffeine and alcohol:** Both affect BP and glucose. Standardise intake or eliminate during testing periods.

**Stress:** Acute stress elevates BP and glucose. Use a daily stress rating (1–10 scale) as a covariate.

### What a positive result would look like

**Blood pressure variability:** A ≥5 mmHg reduction in mean systolic BP AND a ≥3 mmHg reduction in standard deviation when switching from morning to evening dosing (or vice versa). This would suggest that timing matters for your personal physiology.

**Glucose control:** A ≥20% reduction in peak postprandial glucose AND a ≥30-minute reduction in time to peak when consuming a low-glycaemic-index breakfast compared to a high-glycaemic-index breakfast. This would indicate that meal composition meaningfully affects your glucose response.

**Stroke recognition:** Achieving 100% accuracy on FAST symptom identification within 30 seconds (from a cold start, without prompting). This is the minimum threshold for effective bystander response.

### Important caveats for self-experimenters

**Do not test tPA or thrombectomy on yourself.** These are emergency medical interventions that require hospitalisation, imaging, and expert supervision. The guideline's recommendations are for healthcare providers, not individuals.

**Do not use this guideline to self-diagnose or self-treat stroke.** If you suspect a stroke, call emergency services immediately. The FAST acronym is a screening tool, not a diagnostic test.

**Blood pressure and glucose self-experiments are low-risk** if you are under medical supervision. Do not change prescribed medications without consulting your doctor.

**The guideline's evidence is population-level.** Your individual response to any intervention may differ substantially from the averages reported in trials. Use self-experimentation to discover your personal