Part 1: Executive Summary

The guideline authors systematically reviewed evidence on resuscitation interventions including: chest compression depth and rate, use of automated external defibrillators (AEDs), extracorporeal CPR, targeted temperature management after cardiac arrest, drug therapies during resuscitation (epinephrine, amiodarone, lidocaine), airway management strategies, and prognostication tools. Comparators varied by question but generally involved standard care versus alternative approaches. Outcome measures included survival to hospital discharge, neurologically intact survival, return of spontaneous circulation, and quality of CPR delivery.

This is not a single study but a synthesis of evidence from multiple trials. The underlying studies included hundreds of thousands of patients across in-hospital cardiac arrest (IHCA) and out-of-hospital cardiac arrest (OHCA) settings, ranging from neonates to adults. Specific populations varied by PICO question — for example, studies on bystander CPR included community-dwelling adults, while studies on targeted temperature management included comatose post-cardiac arrest adults.

The guideline used the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) framework for evidence quality assessment. Recommendations were classified using the AHA system: Class I (benefit >>> risk), Class IIa (benefit >> risk), Class IIb (benefit ≥ risk), Class III (no benefit or harm). Levels of Evidence ranged from A (high-quality randomised trials) through B-R (randomised), B-NR (non-randomised), C-LD (limited data), to C-EO (expert opinion). Specific measurement instruments varied by outcome — survival was typically measured at hospital discharge or 30 days, neurological outcome using the Cerebral Performance Category scale or modified Rankin Scale.

**Study design:** This is a systematic review and clinical practice guideline, not a single experiment. The AHA partnered with the International Liaison Committee on Resuscitation (ILCOR) to conduct 165 systematic reviews using PICO (Population, Intervention, Comparator, Outcome) format questions. The GRADE process was used for evidence evaluation, with public comment periods incorporated.

**Randomisation and blinding:** Not applicable to the guideline itself. The underlying evidence included randomised controlled trials (some blinded), observational studies, and registry data. The guideline authors rated the quality of each included study and downgraded evidence based on risk of bias, inconsistency, indirectness, imprecision, and publication bias.

**Duration:** The evidence review cycle was five years (2010–2015), though the guideline marks a transition to continuous evidence evaluation. Individual studies within the review ranged from single-episode resuscitation attempts to multi-year registry analyses.

**What this design can and cannot prove:** A systematic review with GRADE methodology can provide high-confidence estimates of treatment effects when high-quality randomised trials exist. It can identify consistent findings across populations and settings. However, it cannot prove causation for observational data, and the recommendations represent expert consensus that may change with new evidence. The guideline explicitly notes that only 1% of recommendations were based on Level of Evidence A (highest quality), highlighting the persistent knowledge gap in resuscitation science.

**Major methodological weaknesses:** The vast majority of recommendations (46%) were based on limited data (LOE C-LD), and 23% were based solely on expert opinion (LOE C-EO). Only 15% of recommendations came from randomised studies. This reflects the inherent difficulty of conducting placebo-controlled trials in emergency resuscitation settings where informed consent is often impossible and randomisation logistically challenging.

**Bystander CPR rates:** Social media dispatch systems increased bystander CPR rates from 48% to 62% (absolute increase 14 percentage points, p-value not reported for this specific comparison in the executive summary)

**Compression depth:** The recommended compression depth was updated to 2–2.4 inches (5–6 cm) for adults, with a rate of 100–120 compressions per minute

**Epinephrine:** Continued recommendation for epinephrine administration during cardiac arrest (Class IIb, LOE B-R for adult cardiac arrest)

**Targeted temperature management:** Continued recommendation for targeted temperature management (32–36°C) for comatose post-cardiac arrest adults (Class I, LOE B-R)

**ECPR (extracorporeal CPR):** New recommendation that ECPR may be considered as a rescue therapy for selected patients with cardiac arrest when conventional CPR is failing (Class IIb, LOE C-LD)

**Public access defibrillation:** Continued recommendation for AED use by laypersons (Class I, LOE C-LD)

**Prognostication:** Updated recommendations for using multiple modalities (clinical examination, neuroimaging, electrophysiology) to predict neurological outcome after cardiac arrest, with specific timing windows (≥72 hours post-arrest for many tests)

**Organ donation:** New information supporting successful transplantation of organs recovered from cardiac arrest victims

**Overall recommendation distribution:** 78 Class I (25%), 217 Class II (68%), 20 Class III (7%); only 3 recommendations (1%) based on Level of Evidence A

This guideline does not report single effect sizes but rather synthesises across multiple studies. For context from the underlying evidence: bystander CPR approximately doubles survival from out-of-hospital cardiac arrest (from ~5% to ~10% in many studies). Targeted temperature management improved neurologically favourable survival by approximately 10–20% relative to no temperature management in the original trials. However, the guideline itself does not provide pooled effect estimates for most recommendations — it provides directional recommendations with confidence ratings.

**Evidence quality:** Only 1% of recommendations based on highest-quality evidence (LOE A); 69% based on limited data or expert opinion

**Generalisability:** Most resuscitation research comes from high-income countries with organised emergency medical services; applicability to low-resource settings is uncertain

**Conflict of interest:** Guideline authors had potential conflicts managed through disclosure and recusal, but industry relationships were present

**Publication bias:** Resuscitation research may over-represent positive findings

**Implementation gap:** Even strong recommendations may not translate into practice change without system-level interventions

**Population heterogeneity:** Cardiac arrest aetiologies vary widely (cardiac vs. non-cardiac, witnessed vs. unwitnessed, shockable vs. non-shockable rhythms), making blanket recommendations challenging

**Temporal changes:** Guidelines were published in 2015; newer evidence may have emerged since publication

**For someone running their own n=1 experiment:**

This guideline is not designed for personal experimentation — it is a clinical practice guideline for healthcare providers managing cardiac arrest. However, the underlying principles can inform personal preparedness:

**What to test:** Your personal ability to perform high-quality CPR. Test compression depth (2–2.4 inches), rate (100–120 per minute), and hand position. Use a CPR feedback device or smartphone app that measures compression metrics.

**Minimum meaningful duration:** Practice for at least 2 minutes continuously (the average time between pulse checks in professional resuscitation). Test your ability to maintain compression depth and rate for 5–10 minutes, which is the typical time until emergency services arrive.

**What to measure:** Compression depth (inches or cm), compression rate (per minute), chest recoil completeness (full release between compressions), and compression fraction (percentage of time compressions are being performed during a resuscitation attempt). Also measure your own heart rate and perceived exertion.

**Key confounds to control for:** Fatigue significantly degrades compression quality within 2 minutes — measure performance in the first minute versus the fifth minute. Surface type (floor vs. bed) affects compression depth. Rescuer body weight and arm position affect mechanics. Practice on a mannequin, not a person.

**What a positive result would look like:** Consistent compression depth of 2–2.4 inches (5–6 cm) maintained for at least 5 minutes, compression rate between 100–120 per minute with less than 10% variation, full chest recoil after each compression, and compression fraction >80% (meaning you spend less than 20% of time not compressing). If using a feedback device, aim for >90% of compressions within target range.

**Important caveat:** CPR is a medical emergency intervention. Do not practice on living people. Use certified CPR mannequins. If you want to be prepared to save a life, take an AHA-certified CPR course — this guideline is the basis for that training. The most impactful personal experiment would be: "Can I learn and retain high-quality CPR skills with a 30-minute practice session every 6 months?" Measure skill retention at 3, 6, and 12 months.

---

# Part 2: Background & Rationale