Skip to main content

Foundational Trials

Kress 2000

Image: Shutterstock / ame435

Publication

Context & Rationale

  • Background
    • Continuous intravenous sedation was widely used to relieve agitation, improve tolerance of an endotracheal tube and facilitate mechanical ventilation. Benzodiazepines or propofol were commonly combined with an opioid.
    • Drug accumulation could prolong unconsciousness after the underlying respiratory illness improved, obscure neurological assessment and delay recognition that ventilation was no longer necessary.
    • Sedation delivery was already becoming a therapeutic target. A 321-patient trial of a nursing-implemented sedation protocol found median ventilation durations of 55.9 versus 117.0 hours with protocol-directed versus conventional sedation, respectively. This established that organisation of sedation care could influence recovery before daily interruption was specifically tested.1
  • Research Question/Hypothesis
    • Would a daily, supervised interruption of sedative and opioid infusions allow earlier awakening and shorten mechanical ventilation and ICU stay compared with interruption at the treating team’s discretion?
  • Why This Matters
    • An apparently necessary supportive treatment could itself be prolonging dependence on organ support. Making sedation a repeatedly reassessed prescription offered a readily reproducible way to reduce avoidable treatment exposure.
    • The intervention also tested whether a bedside assessment of wakefulness could reduce investigations for otherwise unexplained depressed consciousness.

Design & Methods

  • Research Question:
    • Whether daily interruption of continuous sedative infusions would reduce the duration of mechanical ventilation, ICU length of stay and hospital length of stay in mechanically ventilated medical ICU patients.
  • Study Type:
    • Single-centre, randomised controlled trial in the University of Chicago medical ICU, United States.
    • Two-stage allocation: daily interruption versus usual care, followed within each group by random allocation to midazolam or propofol. All four subgroups received morphine for analgesia.
    • Computer-generated assignments were concealed in sealed envelopes. There were 150 randomised patients: 75 per sedation-strategy group.
    • Recruitment dates and funding source: Not reported.
    • The institutional review board waived consent because the intervention fell within the institution’s established standard of care, although it was not routinely applied.
  • Population:
    • Inclusion: adults who were intubated, mechanically ventilated and judged by the ICU team to require continuous intravenous sedation, including patients with agitation or discomfort after the drugs used for intubation had worn off.
    • Exclusions: pregnancy, transfer from another institution where sedatives had already been administered, and admission after resuscitation from cardiac arrest.
    • After randomisation, patients who died or were successfully extubated during the first or second ICU day, before an interruption could occur, were excluded from analysis: 7 intervention and 15 control patients.
    • The analysed population comprised 68 intervention and 60 control patients. It included pulmonary oedema/ARDS, obstructive or other ventilatory failure, sepsis and other medical diagnoses.
  • Intervention:
    • Starting 48 hours after enrolment, an investigator stopped the midazolam or propofol infusion and the morphine infusion simultaneously each day.
    • Infusions remained off until the patient awakened and could follow instructions, or became uncomfortable or agitated and required sedation to resume.
    • Wakefulness required at least three of four actions: opening the eyes to voice, visually following the investigator on request, squeezing a hand on request and protruding the tongue on request.
    • A research nurse observed the patient throughout the interruption and contacted a study physician when the patient awakened; the physician examined the patient and decided whether to resume the infusions.
    • Infusions were restarted at half their previous rates and titrated to the required sedation level. Patients were kept awake only long enough to assess wakefulness; this was not a prescribed prolonged period without sedation.2
    • Sedation was not interrupted during neuromuscular blockade. Daily interruption began after the paralytic drug had been stopped.
  • Comparison:
    • Continuous infusions were adjusted by the usual ICU team, with interruptions at its discretion. Research staff assessed patients daily and recorded drug doses.
    • Both groups used nurse titration to a Ramsay score of 3–4: responding to commands only, or asleep with a brisk response to a light glabellar tap or loud sound.
    • Midazolam: initial intravenous boluses of 0.5–5 mg every 1–5 minutes as required, followed by an infusion starting at 1–2 mg/hour and increased in 1–2 mg/hour steps.
    • Propofol: infusion starting at 5 μg/kg/minute, increased by 5–10 μg/kg/minute every two minutes until adequate sedation.
    • Morphine: initial boluses of 2–10 mg as required and an infusion of 1–5 mg/hour, adjusted using the nurse’s assessment of pain.
    • All care apart from the intervention’s interruption and resumption of infusions remained with the treating ICU team. A standardised ventilator-weaning or extubation algorithm was not reported.
  • Blinding:
    • Sedative agents were open label. Strategy allocation was intended to remain known only to investigators, and specific study endpoints were not disclosed to caregivers.
    • The investigators delivering the interruption could not be blinded. The authors acknowledged that treating clinicians might have recognised allocation; effective blinding of decisions about extubation and discharge was therefore uncertain.
  • Statistics:
    • Power calculation: the effect size sought, α, β/power and required sample size were Not reported.
    • The authors described intention-to-treat analysis, but exclusion of 22 randomised patients means the published comparison was a modified intention-to-treat analysis of the retained 128 patients.
    • Three primary endpoints were specified: ventilation duration, ICU stay and hospital stay. Mortality was an additional recorded outcome.
    • Skewed continuous outcomes were compared with Mann–Whitney U tests and presented as medians with interquartile ranges. Categorical outcomes used χ² tests with Yates’ correction or Fisher’s exact tests.
    • Kaplan–Meier and Cox proportional-hazards analyses assessed time outcomes, with adjustment for age, sex, weight, APACHE II score and type of respiratory failure. Tests were two-sided; multiplicity adjustment was not reported.
  • Follow-Up Period:
    • Patients were followed until hospital discharge. There was no fixed 28-day or 90-day primary endpoint.
    • Long-term cognition, psychological health, functional recovery and quality of life were not assessed in the index trial.

Key Results

No early stopping was reported. The completed study enrolled 150 patients, of whom 128 were analysed after the first-two-day exclusions. A prespecified recruitment target and interim stopping rules were Not reported. Values below are published medians (interquartile ranges), counts or percentages.

Outcome Daily interruption (n=68 analysed) Usual care (n=60 analysed) Effect p value / 95% CI Notes
Duration of mechanical ventilation 4.9 (2.5–8.6) days 7.3 (3.4–16.1) days Adjusted relative risk of extubation 1.9 Median comparison: P=0.004; adjusted estimate: 95% CI 1.3 to 2.7; P<0.001 Primary endpoint. Published difference in medians: 2.4 days. The relative estimate comes from a Cox time-to-event model.
ICU length of stay 6.4 (3.9–12.0) days 9.9 (4.7–17.9) days Adjusted relative risk of ICU discharge 1.6 Median comparison: P=0.02; adjusted estimate: 95% CI 1.1 to 2.3; P=0.02 Primary endpoint. Published difference in medians: 3.5 days.
Hospital length of stay 13.3 (7.3–20.0) days 16.9 (8.5–26.6) days Not reported P=0.19; 95% CI Not reported Primary endpoint; no statistically significant difference.
In-hospital mortality 36.0% 46.7% Not reported P=0.25; 95% CI Not reported Secondary outcome. Percentages reproduced as published; a mortality benefit was not established.
Investigations for altered mental status 6 (9%) 16 (27%) Not reported P=0.02; 95% CI Not reported Intervention: 6 brain CT scans. Control: 13 CT scans, 2 MRI scans and 1 lumbar puncture.
Days with documented wakefulness 85.5% 9.0% Not reported P<0.001; 95% CI Not reported Awake at any time during the day; denominator was days during which sedative infusions were administered.
Device-removal complications 3 patients (4%) 4 patients (7%) Not reported P=0.88; 95% CI Not reported Intervention: 2 self-extubations and 1 central catheter removal. Control: 4 self-extubations.
Reintubation 12 patients 18 patients Not reported P=0.17; 95% CI Not reported No statistically significant difference; the study was small for safety outcomes.
  • Clinical signal:
    • Ventilation and ICU stay were shorter, alongside substantially more days with wakefulness. Hospital stay and mortality did not show statistically significant differences.
    • The Cox estimates describe the relative rate of reaching extubation or discharge over time; they should not be interpreted as binary risk ratios for eventual successful extubation or survival.
  • Sedative subgroups:
    • Among midazolam recipients (37 versus 29 patients), median total midazolam exposure was 229.8 versus 425.5 mg (P=0.05), and morphine exposure was 205 versus 481 mg (P=0.009).
    • Among propofol recipients (31 per group), median total propofol exposure was 15,150 versus 17,588 mg (P=0.54), and morphine exposure was 352 versus 382 mg (P=0.33).
    • Drug-specific clinical effect estimates and formal treatment-by-drug interaction tests were Not reported. These small subgroups cannot establish that the strategy worked only with midazolam or equally well with both agents.
  • Safety and interpretation:
    • Neither intervention-group self-extubation occurred during the interruption itself. Tracheostomy occurred in 12 versus 16 patients (P=0.31).
    • The small number of device events cannot establish equivalent safety. Interpretation of every outcome must also account for the unequal post-randomisation exclusions.

Internal Validity

  • Randomisation and allocation:
    • Computer generation and sealed envelopes support random allocation. Envelope opacity, sequential numbering and safeguards against advance access were Not reported.
  • Post-randomisation exclusions:
    • Of 75 patients allocated per arm, 7 intervention and 15 control patients were removed because of early death or extubation. Outcomes for all 150 randomised patients were not provided.
    • These exclusions condition the comparison on events occurring after allocation. The numbers excluded for death versus rapid extubation were not separated, so the direction of any resulting bias cannot be determined.
  • Performance and detection bias:
    • Daily research-nurse observation and physician examination formed part of intervention delivery. Controls also underwent daily research assessment, but the intensity and clinical consequences of attention were not identical.
    • Extubation, discharge and ordering neurological investigations depend on clinician decisions. Concealing the named endpoints did not ensure that caregivers remained unaware of the sedation strategy.
  • Adherence, separation and crossover:
    • Wakefulness occurred on 85.5% versus 9.0% of infusion days, demonstrating a substantial difference in achieved arousal despite the shared Ramsay target.
    • In 18 of 60 controls, infusions were temporarily stopped before their final day of administration. Across control patients, interruptions occurred on 0–54% of these non-final days.
    • This was permitted control-group exposure to the intervention component, rather than wholesale switching of allocated strategy. Complete denominators for all eligible and missed intervention-day interruptions were Not reported.
  • Baseline characteristics and heterogeneity:
    • Median age was 57 versus 61 years and APACHE II score 20 versus 22. ARDS/pulmonary oedema accounted for 20 versus 15 patients, COPD/ventilatory failure for 22 versus 17, and sepsis for 10 versus 15.
    • Patients had substantial illness severity and a plausible opportunity to benefit from avoiding drug accumulation. The small sample could not reliably test variation in response by diagnosis, illness severity or sedative agent.
  • Timing and intervention dose:
    • Interruption began 48 hours after enrolment; the trial therefore did not test sedation minimisation from the moment of intubation.
    • The intervention combined complete daily interruption, assessment of awakening and restart at half the previous dose. It did not isolate the effect of interruption frequency, duration, opioid withdrawal or dose reduction.
    • Within the intervention arm, average infusion time was 22.8 hours/day with propofol versus 18.7 hours/day with midazolam (P=0.05), illustrating that a scheduled interruption did not produce a uniform period without drug exposure.
  • Key delivery aspects and adjunctive treatment:
    • Seven patients in each group received cisatracurium; sedation was maintained during paralysis. Five patients in each group required non-invasive ventilation after extubation.
    • The shared sedation target and drug protocols reduced treatment variation, but no common weaning algorithm was reported. Staffing and willingness to assess readiness for extubation could influence the observed effect.
  • Outcome assessment and statistical rigour:
    • The primary outcomes were clinically relevant and their recorded times were objective, although the decisions determining those times were not fully protected from bias.
    • Non-parametric comparisons suited the skewed duration data, and adjusted Cox analyses were directionally consistent with the main comparisons.
    • The absence of a reported power calculation prevents assessment of whether the intended precision was achieved. Three primary endpoints were tested without a reported multiplicity strategy.
    • Handling of death in the time-to-extubation and discharge analyses was Not reported. With substantial mortality, interpretation requires attention to death as a competing event; shorter observed treatment duration is not automatically equivalent to more time alive and free of support.

Conclusion on Internal Validity: Moderate. Random allocation, marked separation in wakefulness and consistent ventilation and ICU results support a treatment effect in this setting, but unequal post-randomisation exclusions, uncertain blinding of clinical decisions and incomplete statistical reporting limit confidence in its precise magnitude.

External Validity

  • Population representativeness:
    • The mixed medical ICU population supports relevance across several causes of respiratory failure requiring ongoing sedation. It does not directly represent routine postoperative care, major trauma, primary neurocritical illness or children.
    • Pregnant patients, patients admitted after cardiac arrest and transferred patients already exposed to sedatives were excluded. The analysed cohort also omitted patients dying or recovering rapidly during the first two days.
  • Contemporary sedation practice:
    • The comparator’s 9.0% of days with wakefulness leaves considerable room for improvement. The magnitude of benefit should not be transported unchanged to units already achieving consistently light sedation.
    • Modern care prioritises regular pain assessment, treatment of pain before adding sedation and lighter sedation targets when clinically appropriate. This differs from routinely restarting both hypnotic and morphine infusions towards Ramsay 3–4.3
    • The incremental value of adding interruption to an effective sedation protocol was tested directly in SLEAP, with no demonstrated shortening of time to successful extubation.4
  • Implementation and clinical boundaries:
    • Interruption requires immediate assessment and a prompt response to distress. Delivery by dedicated research staff in one academic ICU does not establish the same effectiveness with different staffing or clinical workflows.
    • The study withheld interruptions during paralysis. It does not establish that interrupting sedation is appropriate when deeper sedation is required for a specific respiratory or neurological indication.
    • No additional drug or device was required, but staff time and monitoring were integral components; resource requirements cannot be inferred solely from reduced drug consumption.

Conclusion on External Validity: Moderate for comparable medical ICU patients receiving continuous sedation, but limited for predicting the additional benefit in contemporary units already delivering reliable light sedation. The principle of repeated reassessment is widely applicable; the original effect size and exact drug-stopping regimen are context dependent.

Strengths & Limitations

  • Strengths:
    • A randomised test of an inexpensive, actionable change to everyday care, with clinically important primary outcomes.
    • A concrete awakening assessment, explicit restart rule and defined exception during neuromuscular blockade made the intervention reproducible.
    • Both groups had the same nominal sedation target, and independent allocation of sedative agent reduced confounding by clinician drug preference.
    • Substantial separation in wakefulness, with coherent reductions in ventilation duration, ICU stay and neurological investigations.
    • The authors explicitly recognised that the low event count limited assessment of adverse effects.
  • Limitations:
    • Single-centre study with 128 analysed patients and 22 exclusions after randomisation.
    • No reported power calculation, separate statistical analysis plan or comprehensive prospective safety assessment.
    • Uncertain caregiver blinding, additional intervention delivery by research staff and no reported standardised weaning algorithm.
    • A comparator characterised by infrequent wakefulness, limiting estimates of benefit over current light-sedation practice.
    • No established survival benefit and no index-trial assessment of delirium, patient-reported distress, long-term cognition or psychological recovery.

Interpretation & Why It Matters

  • Sedation can delay recovery
    • The trial made unnecessary sedative exposure a credible explanation for continued ventilation after physiological improvement. Failure to awaken could no longer be attributed automatically to the underlying illness.
    • The observed 2.4-day reduction in median ventilation duration and 3.5-day reduction in median ICU stay were clinically substantial in the tested setting.
  • Awakening is an assessment
    • A supervised interruption tested the continuing need for sedation and made the patient available for neurological and respiratory assessment. The likely benefit involved both reduced drug accumulation and earlier recognition of recovery; their separate contributions were not measured.
    • Fewer neurological investigations are compatible with less drug-related diagnostic uncertainty. However, 3 of 6 intervention investigations and 4 of 16 control investigations identified an explanation for altered mental status, so reduced testing cannot be equated with elimination of wholly unnecessary investigations.
  • The enduring clinical objective
    • The practical objective is to maintain comfort while preventing unnecessary unconsciousness and repeatedly assessing readiness for liberation from ventilation.
    • Subsequent evidence supports selecting a reliable method of achieving that objective, with daily interruption adding less when an effective light-sedation protocol is already in place.34

Controversies & Other Evidence

  • Was the control group excessively sedated?
    • Wakefulness on only 9.0% of control days raises the central comparator concern. A benefit over that practice does not establish superiority over continuous infusions titrated to a lighter level throughout the day.2
    • Hong and colleagues proposed Ramsay 2–3 rather than 3–4 as an alternative route to reducing sedative exposure. The authors agreed that avoiding oversedation was the goal, but argued that daily interruption enabled awakening while retaining adequate sedation when needed.2
    • The causal distinction matters: increased wakefulness is a plausible mediator of interruption, so it does not invalidate the randomised comparison. It does prevent attributing benefit to a unique property of the scheduled pause rather than the broader reduction in excessive sedation.
    • Heffner warned that routine deep sedation could become further entrenched if a brief daily awakening were treated as sufficient compensation. That warning anticipated the later shift towards lighter sedation across the whole day.5
  • Research attention and the content of the intervention:
    • Riker and Fraser questioned whether specialist research-team attention contributed to better outcomes. The authors clarified that only interruption and resumption were research-team decisions; all other management remained with the usual ICU team.2
    • That clarification reduces concern about wholly different care teams but does not separate the effects of drug cessation, bedside observation, examination and prompting recognition of recovery. The tested intervention was a care process with several components.
    • The reply also clarified that patients were kept awake only long enough to assess them. Describing the protocol as an extended daily period of wakefulness overstates what was delivered.2
  • Excluding early deaths and early extubations:
    • The imbalance of 7 versus 15 exclusions is more consequential than ordinary loss to follow-up: eligibility for the published analysis depended on outcomes after randomisation.
    • Because the intervention had not yet begun, these early outcomes cannot straightforwardly be credited to or blamed on the scheduled interruption. Nevertheless, deleting them can disturb baseline prognostic balance and changes the question from an admission-level treatment policy to a selected cohort still requiring ventilation.
    • The mixed reasons for exclusion point in different prognostic directions. Without their separate counts and an analysis retaining all randomised patients, neither the size nor direction of resulting bias can be reconstructed.
  • Safety, comfort and psychological outcomes:
    • Heffner identified unmeasured distress, recollection, withdrawal syndromes and cardiovascular stress as important omissions. These were concerns about incomplete assessment, not harms demonstrated by the trial.5
    • The correspondence’s claim that device removal had not been assessed was incorrect: the original trial recorded these events, and the authors corrected the point. However, 3 versus 4 events provides little reassurance about uncommon harms.2
    • A later psychological study assessed 32 patients, 13 after daily interruption and 19 controls, at least six months after discharge. Impact of Events scores were 11.2 versus 27.3 (P=0.02), and PTSD diagnoses occurred in 0/13 versus 6/19 (P=0.06).6
    • Only 32 of 105 screened hospital survivors participated, and the cohort combined original-trial participants with contemporaneous patients outside that trial. These findings are reassuring but cannot establish psychological protection or exclude harm in less readily followed survivors.6
    • A prospective observational study of 74 ventilated patients with coronary risk factors found increased haemodynamic and catecholamine responses during interruption, but no increase in the proportion of monitored time with myocardial ischaemia (P=0.17). This addresses the physiological concern without providing definitive evidence about rare cardiac events.7
  • Complications after the index trial:
    • A blinded retrospective review obtained 126 of the original 128 patient charts and identified 13 complications in the interruption group versus 26 in controls (P=0.04), across seven complication categories.8
    • This was an additional analysis of the same cohort, not independent replication. Shorter exposure to ventilation and ICU care could mediate fewer complications; the review does not establish a direct protective effect for each individual complication.8
  • Coordinating awakening with breathing trials:
    • The 336-patient Awakening and Breathing Controlled trial paired daily spontaneous awakening trials with spontaneous breathing trials, compared with usual sedation plus spontaneous breathing trials. Days breathing without assistance to day 28 were 14.7 versus 11.6; mean difference 3.1 days; 95% CI 0.7 to 5.6; P=0.02.9
    • One-year mortality favoured the paired strategy (HR 0.68; 95% CI 0.50 to 0.92; P=0.01), but self-extubations were more frequent: 16 versus 6 (P=0.03). Reintubations after self-extubation were 5 versus 3 (P=0.47).9
    • These findings support coordinated liberation from sedation and ventilation. The combined strategy and longer follow-up do not justify assigning its mortality result to the original Kress intervention alone.9
  • Does interruption add benefit to protocolised light sedation?
    • SLEAP randomised 430 patients in 16 ICUs to a sedation protocol with or without daily interruption. Median time to successful extubation was 7 days in both groups: HR 1.08; 95% CI 0.86 to 1.35; P=0.52.4
    • Interruption increased mean daily midazolam-equivalent exposure (102 versus 82 mg), fentanyl-equivalent exposure (1,780 versus 1,070 μg) and nursing workload scores (4.22 versus 3.80).4
    • The result supports a comparator-dependent interpretation of Kress: the incremental benefit of a scheduled interruption may be small when sedation is already effectively titrated. It does not establish equivalence between all sedation strategies.
  • What the pooled evidence establishes:
    • The Cochrane review included nine trials and 1,282 patients. Daily interruption was associated with a 13% reduction in the geometric mean duration of ventilation, but the 95% CI ranged from a 26% reduction to a 2% increase.10
    • ICU mortality was not clearly different (RR 0.96; 95% CI 0.77 to 1.21), and the estimate for accidental endotracheal tube removal was imprecise (RR 1.07; 95% CI 0.55 to 2.12).10
    • Differences in control sedation, drugs, staffing and liberation protocols limit the meaning of a single pooled treatment effect. The review does not demonstrate a universal reduction in ventilation time or definitively exclude safety differences.10
  • Early sedation and the limits of observational inference:
    • In a 251-patient prospective cohort across 25 Australian and New Zealand hospitals, deeper sedation during the first 48 hours was associated with delayed extubation and higher hospital and 180-day mortality after adjustment.11
    • This supports attention to sedation from the start of critical illness, a period not targeted by the Kress intervention. Residual confounding by illness severity and the indication for sedation prevents interpreting the association as proof that early interruption improves survival.11
  • Less sedation does not imply that no sedation is always better:
    • NONSEDA randomised 710 patients, with 700 in the modified intention-to-treat analysis, to no sedation or light sedation with daily interruption. Ninety-day mortality was 42.4% versus 37.0%; difference 5.4 percentage points; 95% CI −2.2 to 12.2; P=0.65.12
    • Ventilator-free and ICU-free days did not differ significantly. The trial did not establish superiority of nonsedation, reinforcing the need to balance wakefulness with individually assessed comfort and clinical requirements.12
  • Current guidance and contemporary trials:
    • The 2018 PADIS guideline suggests light over deep sedation and states that either daily interruption or nursing protocols can achieve it. That statement does not require both approaches in every patient.3
    • The 2025 focused PADIS update conditionally favours dexmedetomidine over propofol when light sedation and/or reducing delirium are the highest priorities. It addresses drug selection and does not newly establish an independent benefit from daily interruption.13
    • Published subsequently, A2B analysed 1,404 patients in 41 UK ICUs and found no superiority over propofol for time to successful extubation with dexmedetomidine (subdistribution HR 1.09; 95% CI 0.96 to 1.25; P=0.20) or clonidine (1.05; 95% CI 0.95 to 1.17; P=0.34). Severe bradycardia was more frequent with both α₂ agonists.14
    • A2B addresses choice of sedative within contemporary care, rather than daily interruption itself. It cautions against assuming that changing the drug necessarily reproduces the benefit of changing an oversedating care process.14
    • The 2026 Dutch PADIS guideline integrates pain, sedation, delirium, mobility and sleep care, including the organisation of care. It places sedation reduction and awakening/breathing assessments within this wider framework, consistent with treating daily interruption as one component of a patient-centred strategy.15

Summary

  • In a single medical ICU, daily interruption of hypnotic and morphine infusions beginning 48 hours after enrolment reduced ventilation duration and ICU stay compared with interruption at the treating team’s discretion.
  • Median ventilation duration was 4.9 versus 7.3 days (P=0.004), and ICU stay was 6.4 versus 9.9 days (P=0.02); hospital stay and mortality did not differ significantly.
  • Patients were awake on 85.5% versus 9.0% of infusion days, supporting a substantial difference in sedation exposure and assessment opportunity.
  • Interpretation is constrained by 22 post-randomisation exclusions, uncertain blinding of clinical decisions, the infrequently awake control group and limited safety assessment.
  • The lasting lesson is reliable avoidance of unnecessary sedation, with daily interruption selected according to the patient’s condition and the effectiveness of the unit’s existing sedation and liberation practices.

Overall Takeaway

Kress was a landmark demonstration that the delivery of sedation could materially influence the duration of mechanical ventilation and ICU care. Its enduring contribution is the requirement to reassess the need for sedation and expose readiness for recovery; subsequent evidence supports achieving this through reliable, individualised sedation and liberation practices, with daily interruption as one available method.

Overall Summary

  • Daily interruption shortened ventilation and ICU stay in the Kress setting; contemporary practice should preserve its central principle of avoiding unnecessary unconsciousness while ensuring comfort and coordinated assessment for ventilator liberation.

Bibliography

Last updated September 9th, 2026