Publication

• Title: Carbocisteine or Hypertonic Saline for Acute Respiratory Failure
• Acronym: MARCH
• Year: 2026
• Journal published in: New England Journal of Medicine
• Citation: Connolly B, Dickson N, Campbell C, et al; MARCH Trial Investigators. Carbocisteine or hypertonic saline for acute respiratory failure. N Engl J Med. Published online June 10, 2026.

Context & Rationale

• Background

  Invasive mechanical ventilation disrupts normal airway clearance through impaired cough, impaired mucociliary function, altered humidification, endotracheal tube effects, sedation, immobility, and changes in secretion volume and rheology.
  
  Mucoactive agents are therefore commonly used in mechanically ventilated patients with thick or difficult-to-clear secretions, despite major uncertainty about whether they improve clinically important outcomes.
  
  A UK survey reported active mucoactive prescribing in 83% of adult ICUs and point-prevalence use in 411/993 (41.4%) invasively ventilated patients, demonstrating both widespread practice and equipoise.¹
  
  Carbocisteine and nebulised hypertonic saline were selected because they are among the most commonly used mucoactive strategies in UK critical care: carbocisteine as an enteral mucoregulator and hypertonic saline as an inhaled expectorant.
  
  Before MARCH, a systematic review found a small, heterogeneous, high-risk-of-bias evidence base in acute respiratory failure, with no trial evidence supporting or refuting carbocisteine and inconsistent low-quality evidence for hypertonic saline.²
• Research Question/Hypothesis

  MARCH tested whether adding carbocisteine, nebulised hypertonic saline, or both to usual airway-clearance management would shorten the duration of mechanical ventilation in critically ill patients with acute respiratory failure and difficult-to-clear secretions.
  
  The factorial design allowed simultaneous estimation of the main effect of carbocisteine and the main effect of hypertonic saline, while also testing for interaction between the two treatments.
• Why This Matters

  Mucoactives are inexpensive and familiar, but they impose prescription burden, administration time, nebuliser-related workflow, possible ventilator-circuit effects, and drug-specific adverse effects.
  
  If effective, they could plausibly reduce secretion retention, atelectasis, airway obstruction, reintubation, pneumonia, and ventilator duration.
  
  If ineffective, routine use represents low-value care in a population already exposed to high treatment burden.
  
  MARCH therefore addressed a common ICU practice with limited evidential support, using a patient-centred ventilator-duration outcome rather than only surrogate secretion measures.

Design & Methods

• Research Question: In critically ill patients aged ≥16 years receiving invasive mechanical ventilation for acute respiratory failure with difficult-to-clear secretions, do carbocisteine or nebulised hypertonic saline, added to usual airway-clearance management, reduce the duration of mechanical ventilation?
• Study Type: Phase 3, multicentre, pragmatic, open-label, randomised controlled trial with a 2×2 factorial design and internal pilot, conducted in 71 adult ICUs across the United Kingdom.³
• Population:
  • Patients aged ≥16 years.
  • Acute and potentially reversible cause of acute respiratory failure, as determined by the treating physician.
  • Receiving invasive mechanical ventilation through an endotracheal tube or tracheostomy.
  • Expected to remain on invasive mechanical ventilation for at least 48 hours.
  • Difficult-to-clear secretions despite usual airway-clearance management, assessed by the treating clinical team.
  • Key exclusions were pre-existing chronic respiratory disease requiring routine mucoactive use, mucoactive treatment started more than 24 hours before enrolment, known adverse reaction to either study drug, expected treatment withdrawal within 24 hours, known pregnancy, previous MARCH enrolment, declined consent, or clinician judgement that participation was not in the patient’s best interests.
• Intervention:
  • Carbocisteine was administered enterally at 750 mg three times daily for up to 28 days.
  • Nebulised hypertonic saline was administered as 6% or 7% solution, 4 mL four times daily, for up to 28 days.
  • The hypertonic saline concentration was chosen pragmatically according to local availability, reflecting empirical interchangeable use in practice.
  • Patients assigned to the combination group received both carbocisteine and hypertonic saline.
  • Study mucoactives were stopped at the earliest of 28 days, first successful unassisted breathing, ICU discharge, death, discontinuation of active treatment, withdrawal request, study-drug-related serious adverse event, or clinician safety decision.
• Comparison:
  • Usual airway-clearance management alone.
  • Usual care included suctioning, heated humidification, respiratory physiotherapy airway clearance, and isotonic saline at clinician discretion.
  • Respiratory physiotherapy was not protocolised and was delivered according to individual patient assessment and local practice.
  • All other ICU care was determined by treating clinicians.
• Blinding: The trial was open-label. Participants, clinicians, and outcome assessors were aware of treatment assignment. Allocation concealment was preserved until randomisation, but the open-label design creates potential performance and detection bias for process outcomes and clinician-reported adverse events. The primary outcome was relatively objective, although still influenced by clinical decisions around liberation from ventilation and death as a competing event.
• Statistics: A total of 1856 participants was required to detect a 1-day reduction in median duration of mechanical ventilation from 7 to 6 days, corresponding to a hazard ratio of 0.86, with 90% power at a two-sided 5% significance level for each main factorial comparison; the sample size was increased to 1956 to allow 5% loss to follow-up. The analysis followed the intention-to-treat principle using observed primary-outcome data, with Cox proportional-hazards models adjusted for site, age, and APACHE II score. The two main comparisons were any carbocisteine versus no carbocisteine and any hypertonic saline versus no hypertonic saline. Interaction between carbocisteine and hypertonic saline was tested before interpreting the main effects.
• Follow-Up Period: The primary outcome was followed to 60 days after randomisation. Clinical and safety outcomes were measured daily to day 28, first successful unassisted breathing, ICU discharge, or death, whichever occurred first. Mortality and health-related quality of life were planned at 60 days and 6 months, although health-economic and quality-of-life analyses were not reported in the primary manuscript.

Key Results

This trial was not stopped early. MARCH reached its planned sample size, with 1956 participants randomised between February 17, 2022, and April 30, 2025. Follow-up of the final recruited participant was completed on October 31, 2025.

• Neither carbocisteine nor nebulised hypertonic saline shortened mechanical ventilation, reduced reintubation, shortened ICU or hospital stay, or improved mortality.
• The trial demonstrated harm: carbocisteine increased clinically important upper gastrointestinal bleeding, while hypertonic saline increased bronchoconstriction requiring bronchodilator therapy and hypoxaemia during nebulisation.
• Prespecified subgroup analyses and sensitivity analyses did not identify a credible patient phenotype in whom routine carbocisteine or hypertonic saline improved the primary outcome.

Internal Validity

• Randomisation and Allocation: Randomisation used an automated web-based or telephone system, randomly permuted blocks, and stratification by site. The randomisation sequence was concealed from all except the trial statistician. Allocation concealment appears robust.
• Drop out or exclusions: A total of 1956 participants underwent randomisation. In the any-carbocisteine comparison, 26/978 were excluded from the intention-to-treat analysis: 14 lost to primary-outcome follow-up and 12 withdrew consent for data use. In the no-carbocisteine comparison, 27/978 were excluded: 14 lost to follow-up and 13 withdrew consent. In the any-HTS comparison, 29/977 were excluded: 14 lost to follow-up and 15 withdrew consent. In the no-HTS comparison, 24/979 were excluded: 14 lost to follow-up and 10 withdrew consent. Missingness was small and balanced.
• Performance/Detection Bias: The trial was open-label. This is important for clinician-reported safety outcomes, respiratory physiotherapy, antibiotic use, and liberation decisions. The primary outcome was more objective than secretion scores or clinician impressions, but duration of mechanical ventilation can still be influenced by treatment expectations, weaning practices, tracheostomy decisions, and death as a competing event.
• Protocol Adherence: Delivery was good but imperfect. In the factorial comparisons, 948 participants received assigned carbocisteine-containing treatment and 17 did not; 951 received assigned hypertonic-saline-containing treatment and 16 did not. Across comparison groups, treatment duration was approximately 12 days and treatment adherence was approximately 85%.
• Baseline Characteristics: Baseline characteristics were well balanced. Mean age was 56.7 vs 57.5 years in the carbocisteine comparison and 56.6 vs 57.7 years in the HTS comparison. APACHE II scores were 17.2 vs 17.0 and 17.2 vs 17.0, SOFA scores were 9.3 vs 9.4 and 9.4 vs 9.3, ARDS was present in 20.9% vs 21.1% and 21.3% vs 20.7%, and baseline pH was approximately 7.40 in all comparison groups.
• Heterogeneity: Clinical heterogeneity was substantial but appropriate for a pragmatic ICU trial. Participants had a broad range of acute respiratory failure aetiologies, including pulmonary, neurological, postoperative, cardiovascular, sepsis-related, trauma-related, and other causes. This strengthens applicability to routine ICU practice but could dilute a treatment effect in a narrower secretion phenotype.
• Timing: Participants had to be receiving invasive mechanical ventilation and expected to remain ventilated for at least 48 hours. Mucoactive therapy that had started more than 24 hours before enrolment was an exclusion, limiting late enrolment after prolonged prior exposure. The exact time from onset of difficult secretions to randomisation was not reported.
• Dose: Carbocisteine 750 mg three times daily and nebulised HTS 4 mL four times daily were pragmatic, commonly used doses. The trial therefore tested real-world dosing rather than a biomarker-titrated or secretion-rheology-guided regimen. It does not exclude benefit from a different dose, timing, delivery device, or narrowly targeted secretion phenotype, but it does test routine bedside use convincingly.
• Separation of the Variable of Interest: Treatment exposure separated meaningfully. In the four-group CONSORT flow, 473/479 analysed participants in the carbocisteine-alone group, 476/481 in the HTS-alone group, and 475/486 in the combination group received their allocated intervention, compared with no study mucoactive in usual care. Non-trial mucoactive contamination was lower among patients assigned active mucoactive exposure: any carbocisteine 45/965 (4.7%) vs no carbocisteine 157/966 (16.3%); any HTS 41/967 (4.2%) vs no HTS 161/964 (16.7%).
• Key Delivery Aspects: The trial tested mucoactives on top of robust usual airway-clearance care. Respiratory physiotherapy was received by 955/965 (99.0%) vs 951/966 (98.5%) in the carbocisteine comparison and 953/967 (98.6%) vs 953/964 (98.9%) in the HTS comparison. Heated humidification was also common: 78.3% vs 75.7% in the carbocisteine comparison and 77.6% vs 76.4% in the HTS comparison.
• Crossover and contamination: Contamination with non-trial mucoactives occurred more often in the no-active-factor groups. This could bias treatment effects toward no difference, especially if off-protocol mucoactives were effective. However, the strong primary-outcome consistency, lack of mortality or extubation signal, and presence of harms make contamination unlikely to explain the overall result.
• Adjunctive therapy use: Usual airway clearance was high and balanced, including suctioning, humidification, respiratory physiotherapy, and optional isotonic saline. There was no evidence that control-group patients received substantially more airway-clearance adjuncts that plausibly masked a large benefit of the study drugs.
• Outcome Assessment: The primary outcome used a detailed operational definition: time from randomisation to first successful unassisted breathing, maintained for 48 hours, or death. The definition included extracorporeal lung support, invasive ventilation, and non-invasive ventilation with pressure or volume support, and excluded high-flow oxygen and CPAP. This is a clinically meaningful outcome, but death can shorten ventilator duration for an unfavourable reason, which is why reported mortality and sensitivity analyses are important.
• Statistical Rigor: The trial followed a prespecified factorial strategy, tested interaction before main-effect analysis, used intention-to-treat analysis, adjusted for site, age, and APACHE II score, and used observed primary-outcome data because missingness was rare and not differentially distributed. Secondary and subgroup analyses were not adjusted for multiplicity and should be interpreted as exploratory.

Conclusion on Internal Validity: Internal validity is strong for the main clinical conclusion. Randomisation, allocation concealment, large sample size, balanced missingness, prespecified analysis, adequate treatment exposure, and consistent primary and secondary findings support the inference that routine carbocisteine or HTS does not shorten ventilation in this population. Internal validity is more limited for clinician-reported safety and process outcomes because treatment was open-label.

External Validity

• Population Representativeness: The trial enrolled a broad and clinically recognisable ICU population: invasively ventilated patients aged ≥16 years with acute respiratory failure, expected to remain ventilated for at least 48 hours, and judged by the clinical team to have difficult-to-clear secretions.
• Severity: Participants were sufficiently ill to test a clinically important ventilator-duration intervention. Mean APACHE II was approximately 17, mean SOFA approximately 9, ARDS was present in approximately 21%, and day-28 mortality was approximately 22–24%.
• Important Exclusions: Applicability is limited for patients already receiving mucoactives for chronic respiratory disease, patients who had received mucoactives for more than 24 hours before enrolment, patients expected to have treatment withdrawn within 24 hours, pregnant patients, non-intubated patients, patients expected to be ventilated for less than 48 hours, and patients without difficult-to-clear secretions.
• Healthcare Setting: The trial was conducted in UK ICUs with high use of respiratory physiotherapy and usual airway-clearance measures. Translation is strongest to high-resource ICUs with similar airway-clearance infrastructure and ventilator-care practices.
• Applicability: The findings apply directly to routine ICU use of enteral carbocisteine and nebulised 6–7% hypertonic saline for difficult-to-clear secretions in acute respiratory failure. They are less informative for patients with cystic fibrosis, bronchiectasis maintenance indications, chronic sputum disorders requiring established mucoactive therapy, or objective mucus-plugging phenotypes not separately defined in MARCH.
• Resource-Limited Settings: Carbocisteine is inexpensive and enteral, but nebulised HTS requires preparation, delivery equipment, staff time, ventilator-circuit management, and monitoring for hypoxaemia and bronchoconstriction. Given no efficacy signal and clear harms, routine HTS is difficult to justify where monitoring or respiratory-therapy resources are constrained.

Conclusion on External Validity: External validity is strong for adult UK-style ICU practice in mechanically ventilated acute respiratory failure with clinician-identified difficult secretions. Generalisability is weaker for chronic airway disease populations, non-invasive respiratory support, paediatrics, late or already established mucoactive use, and settings with very different airway-clearance delivery.

Strengths & Limitations

• Strengths:
  • Large 1956-participant trial addressing a common ICU practice.
  • Multicentre enrolment across 71 ICUs.
  • Efficient 2×2 factorial design testing two widely used mucoactives simultaneously.
  • Pragmatic dosing and usual-care comparator that reflect real-world practice.
  • Clearly defined primary outcome with 60-day follow-up.
  • Balanced baseline characteristics and low, balanced missing primary-outcome data.
  • Prespecified interaction testing and intention-to-treat analysis.
  • Prespecified safety outcomes that identified clinically relevant harms.
• Limitations:
  • Open-label design, with possible performance and detection bias.
  • Difficult-to-clear secretions were determined by clinical judgement rather than an objective secretion score, rheological measure, or mucus burden biomarker.
  • The primary outcome can be shortened by death, making mortality and sensitivity analyses essential to interpretation.
  • Contamination occurred through non-trial mucoactive use, particularly in patients not assigned the corresponding active factor.
  • Health-related quality-of-life and health-economic outcomes were not reported in the primary manuscript.
  • Only carbocisteine and 6–7% HTS were tested; results do not directly address other agents, devices, dosing regimens, or targeted bronchoscopic strategies.
  • Conduct was limited to UK ICUs, which may have different physiotherapy and airway-clearance practices from other health systems.

Interpretation & Why It Matters

• Clinical meaning

  Routine carbocisteine and routine nebulised hypertonic saline should not be expected to shorten mechanical ventilation in invasively ventilated ICU patients with acute respiratory failure and difficult-to-clear secretions.
• Harm without benefit

  The efficacy signal was absent, while carbocisteine increased clinically important upper gastrointestinal bleeding and HTS increased bronchoconstriction and hypoxaemia during nebulisation.
• Practice impact

  MARCH supports deimplementation of routine mucoactive prescribing for mechanically ventilated acute respiratory failure, while preserving clinician judgement for exceptional situations such as objective mucus plugging or established chronic airway-disease indications.
• Research impact

  The trial shifts future research away from broad empiric mucoactive prescribing and towards better phenotyping of secretion burden, mucus properties, airway obstruction, atelectasis, and patient subgroups in whom targeted therapy might still be rational.

Controversies & Subsequent Evidence

• The pre-trial rationale was pragmatic rather than evidence-rich. Mucoactives were common in UK ICUs, but practice variation and widespread use were not evidence of benefit. The main pre-MARCH evidence synthesis found limited, heterogeneous, low-quality evidence and no direct carbocisteine evidence in acute respiratory failure.¹²
• The “difficult-to-clear secretions” phenotype remains underdefined. MARCH deliberately used clinician judgement to maximise pragmatic relevance, but this creates biological heterogeneity. A patient with tenacious airway mucus, a patient with weak cough, a patient with airway oedema, and a patient with radiographic atelectasis may all be labelled as having difficult secretions, yet may not share the same treatment-responsive mechanism.
• The open-label design particularly affects harms and cointerventions. Bronchoconstriction, hypoxaemia during nebulisation, ventilator-circuit dysfunction, respiratory physiotherapy, and antibiotic use are susceptible to ascertainment and response bias. The primary outcome and mortality are less vulnerable, and these remained consistently unaltered.
• MARCH is consistent with previous ventilated-ICU nebulisation evidence. NEBULAE found that on-demand nebulisation of acetylcysteine or salbutamol was non-inferior to routine nebulisation for ventilator-free days and used fewer nebulisations, supporting a general move away from routine prophylactic nebulised mucoactive therapy in mechanically ventilated ICU patients.⁴
• Evidence from chronic airway disease has also been sobering. In bronchiectasis, a large randomised trial found that hypertonic saline or carbocisteine did not significantly reduce exacerbations over 52 weeks compared with usual care, aligning directionally with MARCH despite major differences in population and treatment goals.⁵
• Observational ICU studies cannot resolve treatment-selection bias. A nationwide cohort study in ventilated pneumonia reported an association between early mucoactive use and lower in-hospital mortality, but such findings remain vulnerable to residual confounding, indication bias, and differences in clinician behaviour. MARCH provides stronger randomised evidence for the broad acute respiratory failure population with difficult secretions.⁶
• Guideline and deimplementation literature already favoured restraint. The AARC pharmacologic airway-clearance guideline did not support routine pharmacologic airway-clearance therapy in many hospitalised non-cystic-fibrosis contexts, and respiratory-care deimplementation work has treated routine nebulised mucoactive prescribing as a low-value practice target.⁷⁸
• Protocol restriction of nebulised 3% saline and N-acetylcysteine has been feasible. A respiratory-therapist-led restriction policy reduced low-value nebulised mucoactive use during mechanical ventilation, providing implementation evidence that deimplementation is achievable when supported by local governance and respiratory therapy leadership.⁹

Summary

• MARCH randomised 1956 mechanically ventilated ICU patients with acute respiratory failure and difficult-to-clear secretions to carbocisteine, hypertonic saline, both, or usual care alone.
• There was no interaction between carbocisteine and hypertonic saline: HR 1.01; 95% CI 0.83 to 1.22; P=0.91.
• Carbocisteine did not reduce duration of mechanical ventilation: 186.1 vs 172.7 hours; adjusted HR 0.96; 95% CI 0.87 to 1.05; P=0.34.
• Hypertonic saline did not reduce duration of mechanical ventilation: 184.5 vs 174.3 hours; adjusted HR 1.00; 95% CI 0.91 to 1.10; P=0.98.
• Both interventions caused harm: carbocisteine increased clinically important upper gastrointestinal bleeding, and HTS increased bronchoconstriction and hypoxaemia during nebulisation.

Overall Takeaway

MARCH is a practice-shaping trial because it tested two common ICU mucoactive strategies at scale in the exact population where clinicians use them: ventilated patients with acute respiratory failure and difficult secretions. The result is clear for routine practice: carbocisteine and nebulised hypertonic saline do not shorten mechanical ventilation and introduce measurable harm.

Bibliography

• 1.Borthwick M, McAuley DF, Warburton J, Anand R, Bradley J, Connolly B, et al. Mucoactive agent use in adult UK critical care units: a survey of health care professionals’ perception, pharmacists’ description of practice, and point prevalence of mucoactive use in invasively mechanically ventilated patients. PeerJ. 2020;8:e8828.
• 2.Anand R, McAuley DF, Blackwood B, Clarke M, Haughey J, O’Neill B, et al. Mucoactive agents for acute respiratory failure in the critically ill: a systematic review and meta-analysis. Thorax. 2020;75:623-631.
• 3.Connolly B, Dickson N, Agus A, Blackwood B, Borthwick M, Bradley J, et al. Effectiveness of mucoactives (carbocisteine and hypertonic saline) in addition to usual airway clearance management with usual airway clearance management alone in acute respiratory failure (MARCH): study protocol for a multi-centre 2×2 factorial, randomised, controlled, open-label, phase 3, pragmatic, clinical and cost-effectiveness trial with internal pilot. NIHR Open Res. 2025;5:30.
• 4.van Meenen DMP, van der Hoeven SM, Binnekade JM, de Borgie CAJM, Merkus MP, Bosch FH, et al. Effect of on-demand vs routine nebulization of acetylcysteine with salbutamol on ventilator-free days in intensive care unit patients receiving invasive ventilation: a randomized clinical trial. JAMA. 2018;319:993-1001.
• 5.Bradley JM, O’Neill B, McAuley DF, Chalmers JD, de Soyza A, Hill AT, et al. Hypertonic saline or carbocisteine in bronchiectasis. N Engl J Med. 2025;393:1565-1577.
• 6.Sasaki A, Nakajima M, Shinozaki T, Sasabuchi Y, Ohbe H, Kaszynski RH, et al. Association between early administration of mucoactive agents and in-hospital mortality in patients with pneumonia requiring mechanical ventilation: a nationwide cohort study. J Intensive Care. 2025;13:57.
• 7.Strickland SL, Rubin BK, Haas CF, Volsko TA, Drescher GS, O’Malley CA. AARC clinical practice guideline: effectiveness of pharmacologic airway clearance therapies in hospitalized patients. Respir Care. 2015;60:1071-1077.
• 8.Fleming K, George JL, Bazelak SJ, Roeske JA, Biggs AD, Landry CM, et al. Optimizing respiratory therapy resources by de-implementing low-value care. Respir Care. 2023;68:559-564.
• 9.Truwit JD, Fleming K, Nanchal RS. Empowering respiratory therapists to restrict nebulized 3% saline and N-acetylcysteine during mechanical ventilation. Respir Care. 2025;70:937-945.