Skip to main content

Foundational Trials

MACMAN

Image: Shutterstock / ChaNaWiT

Publication

  • Title: Video Laryngoscopy vs Direct Laryngoscopy on Successful First-Pass Orotracheal Intubation Among ICU Patients: A Randomized Clinical Trial
  • Acronym: MACMAN
  • Year: 2017
  • Journal published in: JAMA
  • Citation: Lascarrou JB, Boisrame-Helms J, Bailly A, et al; Clinical Research in Intensive Care and Sepsis (CRICS) Group. Video laryngoscopy vs direct laryngoscopy on successful first-pass orotracheal intubation among ICU patients: a randomized clinical trial. JAMA. 2017;317(5):483-493.

Context & Rationale

  • Background
    • Tracheal intubation in the ICU is frequent and physiologically hazardous, with high rates of hypoxaemia, cardiovascular collapse, and cardiac arrest.
    • First-pass success is a plausible, modifiable driver of safety because repeated attempts increase duration of apnoea and airway trauma and may compound haemodynamic instability.
    • Videolaryngoscopy improves glottic visualisation in operating theatres, but effectiveness and safety in ICU intubations were uncertain given different case-mix (hypoxaemia, shock), operator mix, and urgency.
    • Prior ICU data were dominated by observational studies and small trials, with inconsistent signals for first-pass success and limited harms reporting.
  • Research Question/Hypothesis
    • In adults requiring ICU orotracheal intubation, does first-attempt use of a Macintosh-blade videolaryngoscope (McGrath MAC) increase successful first-pass intubation compared with standard Macintosh direct laryngoscopy?
    • Hypothesis underpinning sample size: videolaryngoscopy would increase first-pass success from 65% to 80% (15% absolute improvement).
  • Why This Matters
    • Routine adoption of videolaryngoscopy in critical care has major implications for procurement, training, and airway algorithms.
    • A neutral or harmful effect would shift emphasis to physiology-first airway bundles and targeted device use (anticipated difficulty, supervision), rather than universal deployment.
    • Clarifying whether improved laryngeal view translates into clinically meaningful first-pass success in ICU patients is essential before extrapolating from theatre evidence.

Design & Methods

  • Research Question: In adult ICU patients requiring orotracheal intubation, does first-attempt McGrath MAC videolaryngoscopy (indirect view) improve first-pass success compared with Macintosh direct laryngoscopy?
  • Study Type: Multicentre, parallel-group, investigator-initiated, unblinded randomised clinical trial; 7 adult ICUs in France; May 2015 to January 2016.
  • Population:
    • Setting: adult ICU patients requiring orotracheal intubation, for any indication.
    • Inclusion: adults (≥18 years) requiring orotracheal intubation in ICU.
    • Key exclusions: pregnancy; contraindication to orotracheal intubation; nasotracheal intubation; fibreoptic intubation planned; no time for inclusion/randomisation; previous study inclusion; protected persons (guardianship/curatorship); persons deprived of liberty; no health insurance; refusal.
    • Operator strata: randomisation stratified by centre and operator status (expert vs non-expert).
    • Expert definition: ≥5 years ICU experience, or ≥1 year ICU experience after ≥2 years of anaesthesiology training.
  • Intervention:
    • Device: McGrath MAC videolaryngoscope (Covidien).
    • First-pass technique: indirect view on the screen mandated for the first attempt.
    • Adjuncts on first attempt: no endotracheal tube stylet and no gum elastic bougie.
    • After first-attempt failure: repeat laryngoscopy or alternative airway technique at the operator’s discretion, consistent with local guidelines; subsequent laryngoscopy attempts could use videolaryngoscopy with indirect or direct view.
  • Comparison:
    • Device: standard Macintosh direct laryngoscope.
    • Adjuncts on first attempt: no endotracheal tube stylet and no gum elastic bougie.
    • After first-attempt failure: repeat laryngoscopy or alternative airway technique at the operator’s discretion, consistent with local guidelines.
  • Blinding: Unblinded (operators and bedside staff); primary endpoint objective (capnography-confirmed), but open-label delivery could influence process measures and some complication reporting.
  • Statistics: A total of 370 patients were required to detect a 15% absolute increase in first-pass success (from 65% to 80%) with 90% power (type II error 10%) at a 2-sided 5% significance level; primary analysis was intention-to-treat using a mixed-effects logistic regression model (centre random effect; group and operator status fixed effects); missing primary endpoint counted as failure; secondary outcomes were not adjusted for multiplicity.
  • Follow-Up Period: Peri-intubation outcomes during the procedure and up to 1 hour post-intubation; ICU outcomes followed to day 28 (mortality and other follow-up data reported in the supplement).

Key Results

This trial was not stopped early. Recruitment met the planned sample size (371 randomised; target 370); no interim stopping was reported.

Outcome McGrath MAC videolaryngoscopy Macintosh direct laryngoscopy Effect p value / 95% CI Notes
Successful first-pass orotracheal intubation (primary; intention-to-treat) 126/186 (67.7%) 130/185 (70.3%) Absolute difference −2.5% 95% CI −11.9 to 6.9; P=0.60 Success required waveform capnography for ≥4 breaths.
Successful first-pass orotracheal intubation (per-protocol) 126/183 (68.9%) 130/182 (71.4%) Absolute difference −2.5% 95% CI −12.3 to 6.4; P=0.54 Per-protocol population n=365.
Glottic view: Cormack–Lehane grade I 133/176 (75.6%) 93/177 (52.5%) Absolute difference 23.1% 95% CI 13.3 to 32.7; P<0.001 Better visualisation with videolaryngoscopy did not translate into higher first-pass success.
Time to successful intubation (min; median [IQR]) 3.9 (2.8 to 7.0) 3.8 (2.4 to 6.4) Median difference 0.0 min 95% CI −0.6 to 0.4; P=0.74 Measured from start of anaesthetic induction to capnography confirmation.
Any life-threatening complication (prespecified composite) 24/180 (13.3%) 17/179 (9.5%) Absolute difference 3.8% 95% CI −2.7 to 10.4; P=0.25 Composite included severe hypoxaemia, severe cardiovascular collapse, cardiac arrest, and death.
Severe life-threatening complication (post hoc composite) 17/179 (9.5%) 5/179 (2.8%) Absolute difference 6.7% 95% CI 1.8 to 11.6; P=0.01 Post hoc severity categorisation; no multiplicity adjustment across secondary outcomes.
Day-28 mortality 66/185 (35.7%) 67/184 (36.4%) Absolute difference −0.7% 95% CI −10.5 to 9.1; P=0.88 Follow-up outcome; trial not powered for mortality.
  • Videolaryngoscopy improved glottic view (Cormack–Lehane grade I 75.6% vs 52.5%) but did not improve first-pass success (67.7% vs 70.3%).
  • Sensitivity analysis (intention-to-treat) adjusting for MACOCHA >4 showed no difference in first-pass failure (aOR for videolaryngoscopy 1.10; 95% CI 0.69 to 1.75; P=0.69).
  • In the per-protocol MACOCHA subgroup analysis, first-pass success remained similar in MACOCHA >4 (64.9% videolaryngoscopy vs 67.7% direct; absolute difference 2.9%; 95% CI −19.6 to 25.4; interaction P=0.55).

Internal Validity

  • Randomisation and allocation:
    • Centralised, computer-generated randomisation (software), using permuted blocks of 4 and stratified by centre and operator status (expert vs non-expert).
    • Analytic approach matched the stratified design (mixed-effects logistic regression with centre random effect; group and operator status fixed effects).
  • Dropout or exclusions:
    • 371 patients randomised (186 videolaryngoscopy; 185 direct laryngoscopy).
    • Primary endpoint missing in 5 participants; per protocol, these were counted as first-pass failures.
    • Per-protocol analysis included 365 participants (183 videolaryngoscopy; 182 direct laryngoscopy).
  • Performance and detection bias:
    • Unblinded intervention and bedside outcome recording introduce potential performance bias (operator behaviour) and detection bias for clinician-judgement outcomes.
    • Primary outcome definition was objective and pre-specified (capnography-confirmed first pass).
    • Many complications were defined using physiologic thresholds (e.g., oxygen saturation and blood pressure), limiting subjectivity; some events (e.g., aspiration, airway trauma) remain susceptible to reporting differences.
  • Protocol adherence:
    • Allocated device was used for the first attempt in 183/186 (98.4%) in the videolaryngoscopy group and 182/185 (98.4%) in the direct laryngoscopy group.
    • Induction and preoxygenation were protocol-guided but clinician-selected; key elements were similar between groups (e.g., etomidate 89.1% vs 90.7%; succinylcholine 78.3% vs 75.8%; bag-valve-mask preoxygenation 51.6% vs 52.5%).
    • Mandated technique differences (indirect view for videolaryngoscopy first attempt; no stylet/bougie on first attempt in either group) enhanced between-group separation but may not reflect all contemporary videolaryngoscopy practices.
  • Baseline characteristics:
    • Overall severity was high and balanced (SAPS II mean 58.0 vs 57.7; SOFA median 7 vs 7; PaO2:FiO2 median 95 vs 91).
    • Indications for intubation were similar (acute respiratory failure 65.6% vs 63.8%; coma 23.1% vs 24.3%).
  • Heterogeneity:
    • Seven ICUs with broad indications and variable physiology (shock, hypoxaemia) increase pragmatic relevance but add clinical heterogeneity.
    • Most intubations were performed by non-experts (84.4% vs 83.2%), reflecting typical ICU workforce structures but potentially increasing sensitivity to learning-curve effects.
  • Timing:
    • Randomisation occurred at the time of the intubation procedure, with immediate application of allocated device.
    • Time-to-success was defined from the start of anaesthetic induction, not laryngoscope insertion, which may dilute device-related differences in “laryngoscopy time”.
  • Separation of the variable of interest:
    • Laryngeal view was substantially better with videolaryngoscopy (Cormack–Lehane grade I 75.6% vs 52.5%; P<0.001).
    • Despite improved view, rescue bougie use after first-attempt failure was more common with videolaryngoscopy (12.0% vs 5.5%; P=0.03), suggesting challenges with tube delivery/catheterisation.
    • Time to successful intubation was similar (median 3.9 vs 3.8 minutes; P=0.74), indicating that improved view did not produce procedural efficiency gains in this setting.
  • Outcome assessment:
    • Primary outcome was clearly defined and objectively verified.
    • Complication composites were prespecified, but the “severe life-threatening” composite was post hoc, reducing robustness of inference for that signal.
  • Statistical rigor:
    • Planned sample size achieved (371 randomised vs 370 planned).
    • Primary analysis followed intention-to-treat with prespecified handling of missing primary endpoint (counted as failure).
    • Secondary outcomes were not adjusted for multiplicity, so individual secondary P values should be interpreted as exploratory.

Conclusion on Internal Validity: Overall, internal validity appears moderate-to-strong given robust randomisation, minimal attrition, and an objective primary endpoint; however, open-label delivery and extensive, unadjusted secondary comparisons (including a post hoc “severe” composite) limit confidence in some secondary and harm inferences.

External Validity

  • Population representativeness:
    • High-severity adult ICU population (SAPS II ~58; PaO2:FiO2 ~90–95), broadly reflective of many ICU intubations in high-income systems.
    • Exclusions (e.g., “no time for inclusion/randomisation”, planned fibreoptic/nasotracheal intubation, pregnancy) reduce applicability to crash intubations and selected airway pathologies.
  • Applicability:
    • Device-specific (McGrath MAC) and technique-specific (indirect view mandated on first attempt; no stylet/bougie on first attempt), which may not map to current practice in some ICUs that routinely use stylets or different videolaryngoscope designs.
    • Operator-mix (majority non-expert intubators) enhances relevance to training environments; effects may differ where intubations are performed predominantly by airway experts.
    • Centres were in France with guideline-informed practice (e.g., general anaesthesia with hypnotic and neuromuscular blockade); transferability to systems with different drug availability, staffing, and airway pathways is uncertain.

Conclusion on External Validity: External validity is moderate: findings are most generalisable to routine ICU intubations in resourced centres using Macintosh-blade videolaryngoscopy with mixed-experience operators, but extrapolation to crash scenarios, other device designs, or different adjunct strategies is less secure.

Strengths & Limitations

  • Strengths:
    • Multicentre, pragmatic ICU trial with broad indications and high illness severity.
    • Central randomisation with stratification by centre and operator status.
    • Objective, clinically relevant primary endpoint (capnography-confirmed first-pass success).
    • High protocol delivery for first attempt (≈98% received allocated device) and minimal missingness for the primary endpoint.
  • Limitations:
    • Unblinded design may influence operator behaviour and reporting of some complications.
    • Mandated technique (indirect view on first attempt; no stylet/bougie on first attempt) may not reflect optimised or contemporary videolaryngoscopy practice in all ICUs.
    • Control first-pass success was higher than expected (observed ~70% vs assumed 65%), reducing power to detect smaller benefits than the prespecified 15% absolute increase.
    • Secondary outcomes were not multiplicity-adjusted, and the “severe life-threatening complications” signal was based on a post hoc composite definition.

Interpretation & Why It Matters

  • Clinical practice
    • Routine first-attempt McGrath MAC videolaryngoscopy (as delivered in this trial) did not improve first-pass success compared with Macintosh direct laryngoscopy.
    • Improved visualisation alone is an insufficient surrogate for success in ICU intubation, where tube delivery, apnoea tolerance, and haemodynamic instability are dominant constraints.
  • Mechanistic signal
    • The combination of better view (Cormack–Lehane grade I 75.6% vs 52.5%) and higher post-first-attempt bougie use (12.0% vs 5.5%) supports a “see more, struggle to pass” interpretation for Macintosh-blade videolaryngoscopy when used indirectly without a stylet on first attempt.
    • This underscores that device design and adjunct strategy (stylet shaping, bougie use, external laryngeal manipulation) are integral to effectiveness, not optional add-ons.
  • Implementation & safety
    • The neutral primary outcome and uncertain secondary harm signal reinforce the need to implement videolaryngoscopy alongside physiologically optimised airway bundles, minimised attempts, and structured training/supervision.
    • Device adoption decisions should be made on patient-centred endpoints (first-pass success and complications), not laryngeal view alone.

Controversies & Subsequent Evidence

  • Post hoc severe-complication signal:
    • The higher rate of “severe life-threatening complications” with videolaryngoscopy (9.5% vs 2.8%; P=0.01) was derived from a post hoc severity categorisation, within a framework of multiple unadjusted secondary outcomes, increasing the probability of a chance finding and limiting causal inference.1
  • Correspondence on definitions and technique (what was really being tested):
    • Time-to-success was measured from start of anaesthetic induction rather than from laryngoscope insertion, potentially obscuring device-specific effects on laryngoscopy duration and apnoea exposure.
    • For a Macintosh-blade videolaryngoscope, mandating indirect view for first attempt and prohibiting stylet/bougie on the first attempt was challenged as potentially suboptimal and not reflective of common videolaryngoscopy technique, raising concerns about whether the intervention was “videolaryngoscopy as typically practised” versus a constrained technique.
    • The authors’ reply emphasised a pragmatic comparison with symmetrical adjunct restrictions on the first attempt, and argued that improved view without improved first-pass success reflected real-world constraints in ICU intubation.2345
  • Subsequent RCT evidence shifted the evidential centre of gravity:
    • In the single-centre ICU FELLOW trial, first-pass success did not differ (68.9% vs 65.8%; unadjusted P=0.68), aligning with MACMAN’s neutral primary outcome despite improved view with videolaryngoscopy.6
    • In the larger multicentre DEVICE trial of critically ill adults, videolaryngoscopy improved first-attempt success (85.1% vs 70.8%; absolute risk difference 14.3 points; 95% CI 9.9 to 18.7) with similar severe complications (21.4% vs 20.9%; difference 0.5 points; 95% CI −3.9 to 4.9), implying that context, operator factors, and device/technique implementation materially influence clinical effect.7
  • Meta-analytic synthesis:
    • Systematic reviews generally report higher first-pass success with videolaryngoscopy outside the operating theatre but emphasise heterogeneity across devices, operator experience, and setting, and inconsistent effects on patient-centred outcomes—consistent with MACMAN’s “better view, same first-pass success” phenotype and later trial-to-trial variability.89
  • Guideline trajectory:
    • Recent guidance increasingly recommends videolaryngoscopy availability (and often first-line use) for ICU intubation, but places equal emphasis on training, physiologic optimisation, and minimising attempts—principles that sit alongside MACMAN’s device-specific neutral primary outcome and highlight that “videolaryngoscopy” is an implementation strategy, not a single invariant intervention.10111213

Summary

  • In 371 ICU intubations, first-attempt McGrath MAC videolaryngoscopy did not improve first-pass success versus Macintosh direct laryngoscopy (67.7% vs 70.3%).
  • Videolaryngoscopy markedly improved laryngeal view (Cormack–Lehane grade I 75.6% vs 52.5%) but did not reduce time to successful intubation (median 3.9 vs 3.8 minutes).
  • Overall life-threatening complications were not significantly different; a post hoc “severe” composite was higher with videolaryngoscopy (9.5% vs 2.8%), warranting cautious interpretation.
  • Mortality to day 28 was similar (35.7% vs 36.4%); the trial was not powered for survival endpoints.
  • MACMAN is a landmark ICU airway trial because it demonstrates that improved glottic exposure is not a sufficient surrogate for first-pass success, and that device effects depend on technique and context.

Overall Takeaway

MACMAN is “landmark” because it challenged an intuitive assumption—better glottic view equals better intubation success—in the physiologically high-risk ICU context. It showed that Macintosh-blade videolaryngoscopy (as delivered: indirect first-pass view without a stylet/bougie) did not improve first-pass success versus direct laryngoscopy, while highlighting that device effects are tightly coupled to technique, operator factors, and implementation context.

Overall Summary

  • McGrath MAC videolaryngoscopy improved laryngeal view but did not increase first-pass success in ICU intubations.
  • The post hoc severe-complication signal requires cautious interpretation given multiplicity and post hoc definition.
  • Later larger trials and meta-analyses support that videolaryngoscopy benefits depend on device choice, adjunct strategy, and implementation quality.

Bibliography