2. ICU Discharge Criteria & Step-Down

Quick Recap

Cross-cutting protocol — companion to the Undifferentiated ICU Syndromes system, but addressing the mirror-image moment: the transition OUT of the ICU rather than the first hour in. Applies across all disease-specific protocols in this library; cross-reference the relevant system protocol for diagnosis-specific discharge parameters (e.g., vasopressor-free targets in Septic Shock, weaning criteria in Post-Extubation Failure & Reintubation) — this protocol provides the general framework, safety checklist, and evidence base for the discharge/step-down decision itself.

1. Definition

ICU discharge readiness: a state in which a patient no longer requires ICU-level monitoring or organ support and can be safely managed at a lower level of care (step-down/high-dependency unit or general ward) for a sustained period, typically framed as stability sufficient to be maintained for at least the next 48 hours without a high probability of needing to return to ICU-level care.

Premature discharge: transfer of a patient before this threshold is met, driven by bed-capacity pressure rather than clinical readiness — associated with increased ICU readmission, in-hospital mortality, and cost.

Delayed discharge: retention of a patient beyond clinical necessity — wastes ICU resources, increases nosocomial complication risk, and contributes to ICU bed unavailability for other patients requiring admission.

Discharge readiness is a genuinely under-standardized decision. No single validated, universally adopted set of discharge criteria exists (in contrast to the well-established SOFA/APACHE frameworks for admission severity) — a 2022 European Delphi consensus of 27 intensive care experts across 5 rounds produced a 28-item criteria set as one of the first systematic attempts to formalize this decision, reflecting how much of current practice still relies on individual clinician judgment under bed-pressure conditions.

2. Pathophysiology / Conceptual Framework

Discharge readiness is not simply the inverse of admission severity — it requires assessment across several domains that don't always resolve in parallel:

• Physiological stability: resolution of the acute organ dysfunction that necessitated ICU admission (hemodynamic, respiratory, neurological, metabolic, renal)
• Trajectory, not just a snapshot: a patient can meet momentary physiological thresholds while still being on a deteriorating or unstable trajectory — discharge readiness should reflect a sustained state, not a single favorable data point
• Residual risk from the ICU stay itself: ICU-acquired weakness, delirium, deconditioning, and polypharmacy from the ICU admission independently affect a patient's ability to tolerate ward-level (lower-intensity) monitoring and care, even once the primary organ dysfunction has resolved
• Receiving unit capability: the same patient may be appropriately dischargeable to a step-down/high-dependency unit but not yet ready for a general ward — discharge readiness is relative to the destination's monitoring and staffing capability, not an absolute state
• System and timing factors: evidence consistently shows that when a clinically ready patient is physically transferred (time of day, day of week) independently affects outcome, separate from their physiological readiness — this is a genuinely distinct risk factor from clinical stability itself (Section 22)

3. Immediate Stabilization (ABCDE) — Not Applicable in the Traditional Sense

This protocol addresses a transition-of-care decision rather than an acute stabilization scenario. The ABCDE framework is repurposed here as a discharge-readiness verification checklist across systems, to be worked through deliberately before any transfer order is written:

Airway

Breathing

Circulation

Disability

Exposure

Decision point: meeting these physiological thresholds is necessary but not sufficient — the discharge decision should also explicitly address trajectory (Section 6), residual ICU-stay-related risk (Section 17), and timing (Section 18) before a transfer order is written.

4. Focused History (Discharge-Relevant)

• Original admission diagnosis and whether it has been definitively resolved or only stabilized
• Full list of interventions/organ support weaned during the admission and confirmation each is genuinely no longer required (not just currently withheld)
• Baseline functional status and comparison to current status — informs realistic ward-level care expectations
• Comorbidities that may affect ward-level care tolerance (frailty, baseline cognitive impairment, complex nutritional needs)
• Prior ICU readmission history this admission, if any (a strong risk factor for further readmission)
• Goals of care and code status — confirm current and clearly documented before transfer

5. Comprehensive System-wise Examination (Discharge-Relevant)

• Respiratory: work of breathing, secretion burden, SpO2 trend off escalating support
• Cardiovascular: hemodynamic trend off vasopressors, arrhythmia burden
• Neurological: mental status, delirium status, mobility/strength (ICU-acquired weakness assessment — MRC sum score if feasible)
• Renal: urine output trend, need for ongoing renal replacement therapy
• Nutritional: adequacy of oral/enteral intake, swallow function if relevant
• Skin/musculoskeletal: pressure injury status, mobility level
• Lines/devices: review every indwelling line, catheter, and drain — remove anything no longer necessary before transfer, both for infection prevention and to avoid ward staff managing devices they are less familiar with

POCUS integration: not routinely required for discharge assessment, but may be used selectively to confirm resolution of a specific finding that prompted admission (e.g., confirming resolution of an effusion, reassessing volume status before a borderline discharge decision).

6. Syndrome Identification — Reframed as Trajectory Classification

Rather than identifying an admission syndrome, this protocol asks the clinician to classify the patient's discharge trajectory:

• Clearly ready: stable across all systems, sustained trend, low readmission risk profile
• Borderline/watch: meets physiological thresholds but has risk factors for readmission (Section 8) — consider step-down unit over general ward, or a longer observation window, where available
• Not yet ready: active organ dysfunction or unstable trajectory — do not discharge regardless of bed pressure
• Ready but timing-constrained: clinically ready, but current time/day represents a high-risk discharge window (Section 18) — consider whether transfer can safely wait for a lower-risk window without meaningfully compromising ICU bed availability

7. Differential Diagnosis — Reframed as "Reasons This Patient May Not Actually Be Ready"

Must-not-miss (occult reasons for apparent readiness that are actually premature):

• Recently weaned organ support without adequate observation window (e.g., just extubated, just off vasopressors)
• Unrecognized evolving delirium masking as "calm" in a hypoactive presentation
• Subclinical deterioration not yet reflected in vital signs (early sepsis, occult bleeding)
• Inadequate pain/agitation control masked by sedation rather than genuine stability

Common contributors to inappropriate discharge pressure:

• ICU bed capacity pressure driving discharge timing rather than clinical readiness
• Incomplete handoff communication creating false confidence in receiving team's capability to manage residual risk

System-level factors to exclude before finalizing discharge:

• Receiving unit's actual current staffing/monitoring capability (not just its theoretical capability) — a Delphi-identified criterion distinct from the patient's own status

8. Severity Assessment — Readmission Risk Stratification

Established risk factors for ICU readmission (consistently identified across cohort studies): male sex, increased age, comorbid burden, admission source (ward transfers carry different risk than ED/OR admissions), intensity of organ support required during the ICU stay, and total ICU length of stay.

ICU Discharge Readiness Score models: predictive models developed from large multicenter databases (e.g., eICU Research Institute data, >400 ICUs) to estimate post-discharge death and readmission risk — intended as clinical decision support rather than a rigid gatekeeping threshold; not yet universally implemented but represent a maturing area of risk-stratification tooling relevant to a tech-integration approach.

NEWS2 (National Early Warning Score 2): validated for predicting deterioration on general wards and has been specifically studied for predicting appropriate ICU discharge destination — a useful common-language score between ICU and receiving ward teams, since ward staff are typically already familiar with it for post-discharge monitoring.

Step-down unit destination: observational data show lower adjusted odds of readmission (approximately 36% lower in one large cohort analysis) when a step-down/high-dependency unit is used as the discharge destination rather than direct-to-ward, for patients whose risk profile is intermediate — relevant to destination selection, not only readiness timing.

9. Investigations — Discharge-Confirmatory, Not Diagnostic

Immediate bedside: final vital sign trend review, current organ support requirement confirmation

Routine labs: confirm trend (not just most recent value) of key markers relevant to the admission diagnosis is favorable and stable — avoid discharging on a single reassuring data point without trend confirmation

Repeat frequency: not applicable as an ongoing workup — this section functions as a final confirmatory checkpoint rather than an active diagnostic process

10. Point-of-Care Ultrasound

Selective, confirmatory use only where a specific finding needs resolution verification before discharge (e.g., confirming adequate volume status in a patient recently weaned from vasopressors, confirming resolution of a pleural effusion that prompted admission). Not a routine component of every discharge assessment.

11. Evidence-Based Management — The Discharge Process Itself

Pre-Transfer (Structured Verification)

• Work through the ABCDE discharge checklist (Section 3) explicitly, not as an implicit gestalt judgment
• Apply the trajectory classification (Section 6) — do not discharge a "borderline/watch" patient with the same confidence as a "clearly ready" patient without additional safeguards (step-down destination, earlier follow-up, explicit ward escalation instructions)
• Review readmission risk factors (Section 8) and factor destination selection (step-down vs. general ward) accordingly

Structured Handoff (Evidence-Identified as a Vulnerable Point)

• Verbal and written handoff to the receiving team covering: admission diagnosis and course, residual risk factors, specific parameters that should trigger rapid response/escalation, active management plans (delirium, mobility, nutrition, wound care), code status and goals of care, and outstanding results/follow-up needs
• Explicitly communicate any recently weaned support and the observation window completed, so the receiving team understands what "recently stable" means in this specific patient rather than assuming a longer track record than actually exists

Timing Considerations (Section 18 for full evidence detail)

• Where clinically feasible without compromising ICU bed availability for a more urgent admission, prefer daytime discharge (before ~16:00) over nighttime/out-of-hours discharge
• Where out-of-hours discharge cannot be avoided due to bed pressure, ensure explicit additional safeguards: clear escalation criteria communicated, consider step-down destination over direct-to-ward, ensure rapid response team awareness

Post-Transfer

• Confirm the receiving team has actually received and understood the handoff (closed-loop communication), not just that a note was written
• Institutional-level: track readmissions and out-of-hours discharge rates as quality metrics (Section 16)

12. Organ Support — Confirmation of Genuine Independence

Before discharge, confirm for each organ system that support has been weaned and sustained independence demonstrated, not merely that support is "currently paused":

• Respiratory: off invasive/non-invasive support for an adequate observation window
• Cardiovascular: vasopressor-free for a sustained period
• Renal: no ongoing RRT requirement, or a clear outpatient/ward-appropriate RRT plan if chronic
• Nutritional: adequate oral/enteral intake established, not just tolerated once

13. Disease-Specific Therapy — Not Applicable

This is a process/framework protocol rather than a disease-specific therapeutic one; cross-reference the relevant system protocol (Septic Shock, ARDS, Post-Extubation Failure & Reintubation, etc.) for diagnosis-specific discharge parameters that feed into the general framework above.

14. Consultation Matrix

15. Monitoring Framework — Post-Transfer

• Receiving unit responsibility: NEWS2 or institutional equivalent early warning score tracked per ward protocol post-transfer
• Escalation triggers: any deterioration meeting rapid response/medical emergency team activation criteria — ensure these are explicitly communicated at handoff, not assumed to be generically understood
• ICU-level responsibility (where a step-down/liaison model exists): consider a structured ICU outreach/liaison follow-up visit within 24–48 hours of transfer for higher-risk discharges, particularly out-of-hours or borderline-trajectory patients

16. ICU Bundle Checklist — Discharge Bundle

17. Complications (of Discharge Timing/Process Itself)

Early (within 72 hours of transfer):

• ICU readmission — particularly associated with premature discharge and out-of-hours transfer; readmission within 72 hours is generally considered more likely related to residual index-illness instability than a discrete new event
• Unrecognized deterioration on the ward due to lower monitoring intensity than ICU
• Medication errors/care discontinuity during the transition (a recognized vulnerable point in care transitions generally)

Late:

• Prolonged hospital length of stay from a failed initial discharge requiring ICU readmission
• Increased mortality associated with ICU readmission itself (readmitted patients carry a distinct, elevated mortality risk beyond their original admission severity)
• Post-intensive care syndrome (PICS) sequelae becoming apparent only after ward-level care begins, if not adequately handed off

Prevention: structured discharge checklist use, appropriate destination selection by risk profile, avoidance of avoidable out-of-hours discharge, closed-loop handoff communication

Rescue: rapid response/medical emergency team activation on the ward per explicitly communicated escalation criteria; low threshold for ICU readmission discussion if deterioration occurs, rather than delayed recognition

18. Escalation & De-escalation — Timing-of-Discharge Evidence

This is a distinct evidence base from clinical readiness itself and merits explicit attention:

• A systematic review and meta-analysis of 14 studies (953,312 individuals) found a significantly higher adjusted odds of hospital mortality among patients discharged from ICU at nighttime versus daytime (OR 1.31, 95% CI 1.25–1.38) — consistent, low-heterogeneity finding across multiple health systems (UK, Australia, Canada, USA)
• A large Canadian multi-center cohort (five hospitals, ~19,600 patients) found nighttime discharge (19:00–07:59) was common (17.9% of discharges) and associated with adverse outcomes
• A French multicenter cohort (Outcomerea) found ICU discharge at night specifically (not weekend) was independently associated with increased risk of death (adjusted OR 1.54, 95% CI 1.12–2.11) after controlling for severity — notably, admission timing was NOT similarly associated with risk in the same cohort, suggesting the discharge-timing effect is distinct from a general "after-hours care is worse" phenomenon
• Qualitative/mixed-methods work (the REFLECT study) mapping the discharge process found out-of-hours discharge was common and identified process-level, human-factors contributors rather than purely patient-severity factors — the recommendation from this work: facilitate transfer to the ward before 16:00 where possible, and where unavoidable, implement explicit additional support systems for out-of-hours transfers rather than treating them as routine
• Findings on weekend discharge specifically are less consistent than nighttime findings across the literature, suggesting time-of-day may be the more robust and actionable risk factor to target operationally

Practical implication: this timing effect should be treated as a distinct, addressable risk factor in institutional discharge planning — separate from, and additive to, the physiological readiness assessment in Sections 3–8. A clinically ready patient discharged at 2 AM under bed pressure is not equivalent in risk to the same patient discharged at 11 AM.

19. ICU (Re-)Admission Criteria — The Mirror Question

This protocol's endpoint is transfer out; the corresponding re-escalation question is addressed by the relevant disease-specific protocol should the patient deteriorate post-transfer. Institutionally, track:

• ICU readmission rate as a quality metric (acknowledging it is a controversial metric — Section 22)
• Time-to-readmission (readmission within 72 hours vs. later, as these likely reflect different underlying mechanisms — residual index illness vs. a new discrete event)

20. Documentation & Medicolegal Checklist

• Discharge readiness assessment explicitly documented against the ABCDE checklist (Section 3), not just a brief "stable for transfer" note
• Trajectory classification and rationale for destination selection (step-down vs. general ward) documented
• Readmission risk factors considered and documented
• Discharge timing and rationale documented, including justification if an out-of-hours transfer was necessary despite known risk
• Structured handoff content documented (what was communicated, to whom, confirmation of receipt)
• Escalation criteria communicated to receiving team documented
• Code status/goals of care reconfirmed and documented at transfer
• Family/patient communication regarding the transfer documented

21. Key Guidelines

• European Delphi Consensus on ICU Discharge Readiness Criteria (2022, published BMC Health Services Research): 27 international ICU experts, modified online Delphi over 5 rounds, produced a 28-item consensus criteria set spanning patient-specific stability domains and system/receiving-unit-capability domains — among the first systematic, multi-expert attempts to formalize this decision
• SCCM ICU Admission, Discharge, and Triage (ADT) Guidelines: existing guidance, though noted in recent literature as not having been updated since 2016 — a recognized gap given how much discharge-timing and readmission-risk evidence has accumulated since (Section 22)

22. Landmark Trials / Key Studies

23. Controversies

• ICU readmission rate as a quality metric: explicitly noted in the literature as controversial — a very low readmission rate could reflect excessively conservative (delayed) discharge practice rather than genuinely better care, so this metric should be interpreted alongside length-of-stay and bed-utilization data, not in isolation.
• Lack of a single validated discharge criteria set: unlike admission severity scoring (SOFA, APACHE), no universally adopted discharge readiness instrument exists; the 2022 Delphi consensus is a meaningful step but remains expert-consensus-based rather than outcome-validated at scale, and adoption across institutions is still limited.
• Weekend vs. nighttime discharge risk: findings on weekend-specific discharge risk are notably less consistent across studies than nighttime-specific findings, suggesting these may be distinct phenomena with different underlying mechanisms (e.g., weekend risk may be more confounded by broader staffing pattern differences, while nighttime risk may more specifically reflect the acute transition moment) — institutions should be cautious about assuming these are interchangeable risk factors.
• Outdated formal guidance: the most commonly cited society-level guidance (SCCM ADT guidelines) has not been updated since 2016, predating a substantial portion of the readmission-risk and discharge-timing evidence now available — current practice in many institutions likely runs ahead of formal guideline text on this topic.
• Tension between bed-capacity pressure and discharge-timing safety evidence: the evidence favoring daytime discharge is robust, but ICUs frequently face genuine capacity pressure requiring discharge regardless of time of day to accommodate a more acutely ill incoming patient — this protocol does not resolve that tension but aims to ensure it is navigated explicitly (with compensatory safeguards) rather than by default.

24. References

1. Rojas JC, et al. Readmissions and Death after ICU Discharge: Development and Validation of Two Predictive Models. PLOS One. 2012.
2. van Sluisveld N, et al. Delphi study to derive expert consensus on a set of criteria to evaluate discharge readiness for adult ICU patients to be discharged to a general ward—European perspective. BMC Health Serv Res. 2022.
3. Ouanes I, et al. Readmission to the Intensive Care Unit: Incidence, Risk Factors, Resource Use, and Outcomes. Ann Am Thorac Soc.
4. Santamaria JD, et al. After-hours discharges from intensive care are associated with increased mortality. Med J Aust. 2006.
5. Tiruvoipati R, et al. Association Between Nighttime Discharge from the Intensive Care Unit and Hospital Mortality: A Multi-Center Retrospective Cohort Study. J Intensive Care Med.
6. Hluchy MG, et al. Association between time of discharge from ICU and hospital mortality: a systematic review and meta-analysis. J Crit Care / PMC.
7. Renton J, et al. Mortality associated with timing of admission to and discharge from ICU: a retrospective cohort study (Outcomerea). Crit Care. 2011.
8. Vollam S, et al. Patient Harm and Institutional Avoidability of Out-of-Hours Discharge From Intensive Care: An Analysis Using Mixed Methods (REFLECT study). Crit Care Med.
9. Zaidi H, Bader-El-Den M, McNicholas J. Using the National Early Warning Score (NEWS/NEWS2) in different Intensive Care Units to predict the discharge location of patients. BMC Public Health. 2019.
10. Ethical considerations in evaluating discharge readiness from the intensive care unit. [journal citation per source].
11. The Washington Manual of Critical Care, 4th ed. 2025 — comprehensive ICU care chapter.
12. ICU Protocols: A Step-wise Approach, 2nd ed. — comprehensive ICU care chapter.