Skip to main content

Foundational Trials

ABC

Image: Shutterstock / ChaNaWiT

Publication

  • Title: Efficacy and safety of a paired sedation and ventilator weaning protocol for mechanically ventilated patients in intensive care (Awakening and Breathing Controlled trial): a randomised controlled trial
  • Acronym: ABC (Awakening and Breathing Controlled)
  • Year: 2008
  • Journal published in: The Lancet
  • Citation: Girard TD, Kress JP, Fuchs BD, Thomason JWW, Schweickert WD, Pun BT, et al. Efficacy and safety of a paired sedation and ventilator weaning protocol for mechanically ventilated patients in intensive care (Awakening and Breathing Controlled trial): a randomised controlled trial. Lancet. 2008;371:126-34.

Context & Rationale

  • Background
    • Continuous sedation is common during invasive ventilation, yet excessive or prolonged sedation can delay weaning, increase ICU complications, and prolong ICU and hospital stay.
    • Trials prior to ABC suggested that structured ventilator-weaning processes (e.g., protocolised SBTs) and daily interruption of sedatives can shorten ventilation duration; however, these processes were often implemented separately rather than as an integrated liberation strategy.
    • Delirium and coma had emerged as clinically important acute brain dysfunction syndromes linked to prolonged mechanical ventilation and worse long-term outcomes, creating a rationale for sedation-minimising, wakefulness-promoting approaches that might also facilitate earlier extubation.
  • Research Question/Hypothesis
    • Whether a daily paired protocol of spontaneous awakening (SAT) followed immediately by a spontaneous breathing trial (SBT) increases ventilator-free days (and improves downstream outcomes including ICU/hospital length of stay and mortality) compared with usual care in mechanically ventilated ICU patients.
    • Hypothesis: coupling sedation interruption with ventilator liberation testing would create stronger “separation” in wakefulness/ventilatory readiness, leading to faster liberation without excess harm.
  • Why This Matters
    • Mechanical ventilation duration is a causal driver of ventilator-associated complications and ICU resource use; a scalable, protocol-based approach that shortens ventilation could shift population-level outcomes.
    • The intervention targets a modifiable process of care (sedation + weaning coordination), making it attractive for quality improvement and bundle-based implementation.
    • Demonstration of long-term survival benefit (if real) would elevate the intervention beyond a simple efficiency strategy to a patient-centred outcome intervention.

Design & Methods

  • Research Question: In invasively ventilated adult ICU patients, does a daily paired SAT+SBT protocol (vs usual sedation practices with daily SBT protocol) increase ventilator-free days during the first 28 days and improve clinically important outcomes including mortality?
  • Study Type: Multicentre, parallel-group, randomised controlled trial; four tertiary-care hospitals in Tennessee (medical/surgical ICUs); investigator-initiated; open-label (protocolised process-of-care intervention delivered by bedside clinicians).
  • Population:
    • Inclusion: Adults (≥18 years); receiving invasive mechanical ventilation for ≥12 hours; expected to require ventilation for ≥24 additional hours; receiving continuous IV sedation and/or analgesia.
    • Key exclusions: Moribund/expected death within 24 hours; ongoing myocardial ischaemia; neuromuscular blockade use; active seizures or escalating sedation for seizures/agitation; increased intracranial pressure; other protocol-specified safety exclusions.
    • Trial flow: 1,658 eligible; 1,322 excluded (e.g., 322 refused, 36 withdrew consent before randomisation); 336 randomised (168 per group).
  • Intervention:
    • Daily SAT safety screen: If no exclusion criteria (e.g., no active seizures; no escalating sedative dose requirements; no neuromuscular blockade; no active myocardial ischaemia; no increased intracranial pressure), all sedative and analgesic infusions were stopped.
    • SAT failure criteria: Sustained agitation/anxiety (e.g., RASS ≥ +2), sustained respiratory rate >35/min, oxygen saturation <88% (≥5 minutes), acute arrhythmia, or other safety events prompting restart of infusions; if failed, infusions restarted at ~50% of prior dose and titrated.
    • Paired SBT: If SAT successful, patient immediately underwent SBT safety screen (e.g., FiO2 ≤0.50; PEEP ≤8 cm H2O; no vasopressor infusion or low-dose support; no myocardial ischaemia; able to initiate inspiratory efforts).
    • SBT conduct: Trial using low-level pressure support/CPAP; failure criteria included tachypnoea, hypoxaemia, marked distress, or haemodynamic instability; if passed, clinicians were encouraged to proceed with extubation when appropriate.
  • Comparison:
    • Usual care sedation: Sedation/analgesia management at clinicians’ discretion (no mandated daily interruption), reflecting local practice patterns.
    • Daily SBT protocol: Both groups used a daily SBT protocol, but in the control group sedatives were not systematically stopped before SBT (i.e., SAT and SBT were not paired as an integrated sequence).
    • Co-interventions: Other ICU care, including decisions around extubation, tracheostomy, and post-extubation respiratory support, was clinician-directed.
  • Blinding: Unblinded (bedside clinicians and outcome assessors could not be blinded to a wake-up/breathe protocol); primary outcome included objective components (ventilator status) but delirium/coma assessments were potentially susceptible to performance/detection effects.
  • Statistics: Sample size 334 (increased to 336) to detect a 3-day difference in ventilator-free days (assumed mean 12.9 days; SD 10.4), with 80% power at two-sided α=0.05; primary analysis intention-to-treat; time-to-event outcomes analysed with Kaplan–Meier methods and log-rank testing; effect estimates reported as differences with 95% CIs where available.
  • Follow-Up Period: Primary endpoint assessed through day 28; survival followed through one year.

Key Results

This trial was not stopped early. Recruitment and follow-up proceeded to the planned sample size; protocol deviations and withdrawals were reported but no early efficacy stopping was reported.

Outcome SAT+SBT (paired protocol) Control (usual sedation + SBT) Effect p value / 95% CI Notes
Ventilator-free days (days 1–28) 14.7 (0.9) 11.6 (0.9) Difference 3.1 95% CI 0.7 to 5.6; P=0.02 Primary outcome; reported as mean (dispersion as published).
Ventilator days 7.4 (0.6) 10.3 (0.6) Difference −2.9 95% CI −4.2 to −1.6; P=0.02 Supports greater separation in liberation trajectory.
Time to discharge alive from ICU (days) 9.1 (5.1–17.8) 12.9 (7.2–24.6) Not reported P=0.01 Reported as median (IQR) time-to-event.
Time to discharge alive from hospital (days) 14.9 (8.3–33.2) 19.2 (11.0–39.5) Not reported P=0.04 Reported as median (IQR) time-to-event.
Duration of coma (days) 0 (0–3) 1 (0–4) Not reported P=0.04 Median (IQR).
Duration of delirium (days) 2 (0–6) 4 (0–10) Not reported P=0.02 Median (IQR); delirium assessed with CAM-ICU as published.
Hospital mortality 29.3% 41.3% Not reported P=0.03 Effect estimate (RR/HR) not reported for hospital mortality.
Mortality up to 1 year 44.3% 58.3% HR 0.68 95% CI 0.50 to 0.92; P=0.01 Number needed to treat to prevent one death: 7 (as reported).
Self-extubation 10.0% 4.0% Difference 6.0% 95% CI 0.6 to 11.8; P=0.03 Higher unplanned extubations with paired protocol.
Reintubation 13.6% 12.3% Difference 1.3% 95% CI −5.9 to 8.6; P=0.73 Despite more self-extubations, overall reintubation was similar.
  • Paired SAT+SBT increased ventilator-free days by 3.1 days (95% CI 0.7 to 5.6) and shortened ICU and hospital time-to-discharge (P=0.01 and P=0.04, respectively) without an increase in overall reintubation.
  • Acute brain dysfunction metrics improved (coma 0 vs 1 day; delirium 2 vs 4 days), aligning with a mechanistic pathway of less sedative exposure and more wakefulness.
  • Survival benefit was most evident at one year (HR 0.68; 95% CI 0.50 to 0.92; P=0.01), while 28-day mortality did not differ significantly (28.0% vs 35.0%; P=0.21).

Internal Validity

  • Randomisation and Allocation:
    • Computer-generated, stratified randomisation with permuted blocks; allocation via sequentially numbered, sealed opaque envelopes.
    • Randomisation after consent and enrolment; group sizes balanced (168 per arm).
  • Drop out or exclusions:
    • Selection before randomisation was substantial (1,658 eligible; 1,322 excluded; 322 refused; 36 withdrew consent before randomisation).
    • After randomisation: intervention arm 1 patient withdrew early and did not initiate protocol (167 analysed); control arm analysed 168 (with 1–2 patients reported as lost to follow-up in the trial profile).
    • Post-randomisation exclusions were few, but the single early withdrawal in the intervention arm potentially makes “pure ITT” less complete.
  • Performance/Detection Bias:
    • Blinding was not feasible; clinicians knew allocation and could alter co-interventions (e.g., readiness-to-extubate decisions, mobilisation practices).
    • Primary outcome (ventilator-free days) is partly objective, though extubation timing has clinician-discretion components.
    • Coma/delirium outcomes are potentially more susceptible to assessment bias in an unblinded care pathway.
  • Protocol Adherence:
    • Clear “separation” was achieved: patients underwent SAT in 90% of opportunities in intervention vs 0% in control; sedatives were held before any SBT in 90% vs 31%.
    • Intervention reduced cumulative sedative/analgesic exposure (e.g., lorazepam 20 vs 39 mg; propofol 366 vs 772 mg; fentanyl 7,500 vs 12,483 μg; all P≤0.02 as published).
    • Control arm had protocol discontinuations (7 discontinued), reflecting real-world feasibility but increasing treatment heterogeneity.
  • Baseline Characteristics:
    • Baseline demographics and severity of illness were broadly balanced (e.g., mean age 58 vs 61 years; APACHE II 27 vs 28; similar pre-enrolment ventilation hours).
    • Case-mix included a substantial sepsis burden and both medical and surgical ICU admissions, supporting clinical relevance.
  • Heterogeneity:
    • Multicentre delivery across four hospitals increases between-site heterogeneity but also reduces centre idiosyncrasy risk.
    • Both arms received an SBT protocol, isolating the added value of daily SAT “pairing” rather than comparing protocolised vs non-protocolised weaning.
  • Timing:
    • Daily screening and intervention delivery occurred early during the ventilated course (eligibility required >12 hours ventilation and anticipated ongoing ventilation), aligning the intervention with the modifiable phase of sedation and weaning decisions.
  • Dose:
    • Intervention “dose” was a daily attempt at complete interruption of sedative/analgesic infusions when safe; the achieved reduction in sedative exposure suggests adequate dosing intensity.
  • Separation of the Variable of Interest:
    • Sedative interruption before SBT: 90% (intervention) vs 31% (control).
    • Cumulative sedative/analgesic dosing differences (published totals): lorazepam 20 vs 39 mg; propofol 366 vs 772 mg; fentanyl 7,500 vs 12,483 μg.
    • Wakefulness at first successful SBT: RASS 0.6 vs −0.1 (P=0.02).
  • Key Delivery Aspects:
    • The protocol operationalised multidisciplinary coordination (nursing-led SAT + respiratory therapy-led SBT), a core implementation feature rather than an ancillary detail.
    • Control care included daily SBT, raising the bar for incremental benefit and reducing the likelihood that results simply reflect “having a weaning protocol at all”.
  • Outcome Assessment:
    • Ventilator-free days and mortality are clinically important and relatively robust outcomes.
    • Delirium/coma metrics were protocol-defined and assessed with structured tools (as published), but the lack of blinding remains a methodological constraint.
  • Statistical Rigor:
    • Sample size target met (336 randomised), and primary endpoint showed a statistically significant difference with a reported 95% CI.
    • Time-to-event endpoints analysed using survival methods; the 1-year mortality effect was reported as a hazard ratio with confidence interval.

Conclusion on Internal Validity: Overall, internal validity appears moderate-to-strong given robust randomisation/allocation concealment and clear intervention–control separation, tempered by unavoidable lack of blinding, meaningful pre-randomisation exclusions/refusals, and minor post-randomisation withdrawal/protocol discontinuation.

External Validity

  • Population Representativeness:
    • Represents a broad adult mixed ICU cohort requiring ongoing mechanical ventilation and receiving continuous sedation/analgesia infusions.
    • Key exclusions (e.g., intracranial hypertension, active myocardial ischaemia, neuromuscular blockade) limit applicability to neurocritical care and some high-risk cardiopulmonary subgroups.
    • High exclusion/refusal rates imply that participants may be a selected subset with greater equipoise, greater protocol feasibility, or differing baseline risk.
  • Applicability:
    • Implementation requires reliable staffing, training, and culture supporting daily SAT and SBT coordination; this may translate well to resource-rich ICUs with established protocols.
    • Settings with different sedation practices (e.g., already light sedation with daily sedation targets) may see attenuated marginal benefit.
    • Resource-limited settings may face barriers (e.g., fewer respiratory therapists, less frequent structured delirium assessments) that affect fidelity.

Conclusion on External Validity: Generalisability is reasonable for adult mixed ICUs with capacity for multidisciplinary protocol delivery, but applicability is more limited in neurocritical care, highly unstable patients excluded by safety screens, and environments lacking staffing/structure to deliver daily paired wake-up and breathing trials.

Strengths & Limitations

  • Strengths:
    • Pragmatic, protocol-based intervention targeting a high-impact process of care.
    • Multicentre delivery with clear intervention–control separation in sedation practice.
    • Clinically meaningful outcomes, including ventilator-free days and 1-year mortality with reported hazard ratio and confidence interval.
    • Control arm received daily SBT protocol, strengthening the inference that benefit derives from pairing SAT to SBT rather than generic weaning protocolisation.
  • Limitations:
    • Unblinded care pathway with potential performance and detection bias, particularly for delirium/coma outcomes.
    • Large pre-randomisation exclusions/refusals, raising concern for selection and implementation feasibility in routine practice.
    • Protocol deviations in control group (discontinuations) and a small post-randomisation withdrawal in the intervention group.
    • Increased self-extubation risk, highlighting the need for careful implementation and monitoring.

Interpretation & Why It Matters

  • Integrated liberation strategy
    • ABC operationalised a “linked” liberation pathway: waking patients (SAT) is treated as a prerequisite test enabling a more informative breathing trial (SBT) and extubation decision-making.
    • The observed improvement in ventilator-free days and shorter ICU/hospital time-to-discharge support the hypothesis that coordination (not merely presence of either protocol alone) matters.
  • Acute brain dysfunction as a target
    • Reduced coma (0 vs 1 day) and delirium duration (2 vs 4 days) provide biologic plausibility for downstream benefits and embed delirium into the liberation narrative as an outcome and mediator.
  • Patient-centred endpoints
    • One-year survival benefit (HR 0.68; 95% CI 0.50 to 0.92) is a striking finding for a process-of-care intervention and underpins ABC’s “landmark” status.
    • However, interpretability depends on replication context, changes in sedation practice over time, and mechanistic credibility given the lack of early mortality signal.

Controversies & Subsequent Evidence

  • The accompanying Lancet commentary framed “wake up and breathe” as a compelling integrated concept while emphasising that safety and feasibility depend on unit culture, staffing, and careful management of agitation/self-extubation risk.1
  • ABC’s lack of blinding raises concern for co-intervention effects (e.g., clinician behaviour around extubation, mobilisation, and sedative choice) that could plausibly influence time-to-extubation and ICU discharge, although the magnitude of sedation separation suggests the intervention itself materially changed exposure.
  • Subsequent randomised evidence suggested that in ICUs already using protocolised sedation, adding a daily sedation interruption may not further reduce ventilation duration and may increase workload or adverse events, challenging the generalisability of ABC’s “SAT” component as a universal add-on.2
  • Trials of lighter/no-sedation strategies reinforced the principle that minimising sedation can shorten ventilation and ICU stay but may trade off with agitation and safety events, underscoring that “less sedation” requires structured delivery rather than simple dose reduction.3
  • Contemporary guidelines incorporate the liberation philosophy: emphasising light sedation, routine assessment, and structured ventilator liberation approaches, while recognising that specific elements (e.g., daily sedation interruption) should be implemented with attention to baseline sedation practice and feasibility.456

Summary

  • ABC tested a daily paired spontaneous awakening + spontaneous breathing protocol in 336 mechanically ventilated adult ICU patients across four tertiary hospitals.
  • The paired protocol increased ventilator-free days (14.7 vs 11.6; difference 3.1; 95% CI 0.7 to 5.6; P=0.02) and reduced ICU/hospital time-to-discharge (P=0.01 and P=0.04).
  • Acute brain dysfunction decreased (coma 0 vs 1 day; P=0.04; delirium 2 vs 4 days; P=0.02), consistent with reduced sedative exposure and greater wakefulness at SBT.
  • Hospital mortality was lower (29.3% vs 41.3%; P=0.03) and 1-year mortality improved (HR 0.68; 95% CI 0.50 to 0.92; P=0.01), while 28-day mortality was not significantly different (28.0% vs 35.0%; P=0.21).
  • Self-extubation was more common (10.0% vs 4.0%; difference 6.0%; 95% CI 0.6 to 11.8; P=0.03), but overall reintubation rates were similar.

Further Reading

Other Trials

Systematic Review & Meta Analysis

Observational Studies

Guidelines

Notes

  • Implementation fidelity is central: ABC is best understood as a coordinated multidisciplinary workflow (SAT immediately informing SBT), not as two independent checklists.
  • Monitor safety signals when implementing (e.g., agitation, self-extubation), and ensure clear criteria for SAT/SBT failure and prompt sedation re-titration.

Overall Takeaway

ABC reframed ventilator liberation as an integrated process: daily awakening is not merely a sedation strategy, but a diagnostic step enabling a more meaningful breathing trial and extubation decision. The trial’s combination of improved ventilator-free days, shorter ICU/hospital stays, and a reported one-year survival benefit made it a cornerstone for modern ICU liberation bundles, while later evidence emphasised that benefit depends on baseline sedation practice, protocol fidelity, and safe implementation.

Overall Summary

  • Daily paired SAT+SBT increased ventilator-free days (difference 3.1; 95% CI 0.7 to 5.6) and reduced ICU/hospital time-to-discharge, with a reported 1-year survival benefit (HR 0.68).
  • Protocol delivery achieved strong separation in sedation exposure (SAT 90% vs 0%; sedatives held before any SBT 90% vs 31%), supporting a plausible mechanism for faster liberation and less delirium/coma.
  • Implementation requires attention to safety (self-extubation higher) and multidisciplinary workflow; later trials suggest daily sedation interruption may add less in settings already achieving light, protocolised sedation.

Bibliography