Providing information for discharge from the ICU

"Providing information for discharge", one of the topics from Section 2.2.1 in the second edition of the CICM Syllabus for the Second Part Examination, is probably a part of a wider ICU discharge management subject matter that is too big for the Communication and collaboration section. Certainly other ICU discharge SAQs such as Question 4 from the first paper of 2025 were mostly in the Section 2.3.1 (Safety and quality) domain. One might argue that Question 6 from the second paper of 2025 was similarly a quality/safety question, as it dealt with discharge directly home from the ICU, but it ended up being smeared across no less than four syllabus topic sections, and each of them (Decision making, Structure and Process, Populations requiring special considerations) each seems so reasonable that there was no way to choose between them. This chapter ended up planted here mostly because the author came to the conclusion that fundamentally, the direct-to-home ICU discharge is an act of communication and trust, where the intensivist throws a ball and hopes that community healthcare teams will catch it.

The expected knowledge from Section 2.2.1 does not apply directly to this topic, because the expectatiuons fit so poorly that any attempt to force the material under these headings would result in herniation. "For the above topics candidates should be able to discuss and demonstrate effective, respectful and empathetic, professional, culturally safe and patient/family centred communication skills" does not apply well to this, ultimately, very bureaucratoprocedural subject. What follows has been structured according to the incompletely understood themes from the literature.

Definition of discharge directly home from the ICU

"Discharge home" could be held to mean many things, but for the purpose of the discussion here, we will ignore "discharged to a home", i.e. nursing home, or "discharged into a fully modified home to remain under the care of a fully supported family who are accustomed to caring for a long-term ventilator dependent patient". Additionally, we will exclude those happy few for whom the Saturday morning after a wild night has resulted in another episode of waking up to a self-extubation,  only to discover that ED have cut up their pants again. They who discharge against medical advice to perform the walk of shame back to the train station were not discharged from the ICU because of a deliberate decision by the medical staff there, and therefore there is nothing further to be learned from their experience. The specific topic of discussion is a decision to discharge a patient from ICU-level care in a hospital to community-level care at home, which represents a significant downward step in the level of medical attention.

That sounds like it should be easy to define and measure, and indeed, all Australian ICUs dutifully complete and submit the ANZICS APACHE data collection forms to the CORE registry, where ICU discharge destination can be reported as "Home"; but no publication currently seems to report any analysis of this, perhaps because the registry unhelpfully flattens "Home", "Ward" and "Other" into the stored integer "3", obliterating the depth of this rich data.  Sure, you could reconstruct it somehow, and for a college that demands research projects from their exiting trainees a widespread neglect of this freely available resource seems unlikely (so surely somebody's project somewhere has already looked at this), but the author freely admits that he did not survey them all for their unpublished works. 

Epidemiology of discharge directly home from the ICU

 In Australia and New Zealand, the publicly visible contemporary denominator is the massive Intensive Care Foundation ICU statistics page, drawing on ANZICS registries (226 ICUs,  2,655 ICU beds across Australia and New Zealand, 200,000 adult ICU admissions), but it contains no metric to describe this outcome, perhaps because the ANZICS data registry mixes "home" with "ward".  The main  data source for local Australian information seems to be the paediatric paper by Kennedy & Numa (2020), who reported an increase in direct discharge home between two service epochs, from 3.7% to 9.5% of PICU survivors, with direct discharge associated with discharge delay. This is obviuously nothing like the aduilt ICU data set (eg. authors observed that bronchiolitis was overrepresented (18%) in this category, which they attributed to the reluctance of the ward teams to use high flow nasal oxygen). Beyond this, there does not appear to be any local conversation about this subject, let alone data. 

Worldwide adult data are larger but still mostly observational and poorly generaliseable because of some choice on the part of the authors to limit their audit to the small local boutiques. For example  Chawla et al (2012) found direct home discharge in only 3.6% of ICU discharges, and then was able to like this to a very high rate of 30-day readmission, but this population consisted of exclusively oncology patients. Lau et al (2018) reported a Canadian experience where the rate increased from 3.1% to 12%, which was  inversely associated with ICU occupancy (i.e. the more the ICU was full, the fewer patents were discharged directly home). Another study followed this up with a deliberate effort to send patients home directly, which was  apparently beloved by all except the involved clinicians.

Canada generally seems to be the most comprehensive source for these data. Of these the largest has to be the Ontario cohort from Martin et al (2020), with 1,222,708 patients, dwarfing Stelfox et al (2018) with a mere 6732 ICU patients. In the former and latter, 21.1% and 14% were discharged directly home, respectively. For the Stelfox cohort, there was no significant increase in 30-day hospital readmission, 30-day ED visits or 1-yr mortality was discovered, which suggests that for this population the discharge decisions were made well.  Case selection is probably doing most of the work, the authors suggested; all the patients were young, low acuity, and with single organ problems. 

Advantages and disadvantages of bypassing ward discharge

In their comments to  Question 6 from the second paper of 2025, the examiners had remarked that "those candidates who incorporated this structure into their answer were rewarded for their attention to the glossary of terms instructions with higher scores", where "this structure" refers to the subheadings suggested by the "discuss" glossary term ("Discuss: Explain the underlying key principles. Where appropriate, this may include controversies and/or advantages and disadvantages"). From this, it follows that they wanted something like:

Advantages	Disadvantages
Shorter hospital stay: avoids a pointless ward admission for a patient who is already fit to leave hospital, which was one of the main findings in Stelfox et al (2018).	Premature discharge risk: bypassing the safety buffer of ward admission does not allow enough time for the team to identify patients who are unready for home discharge for reasons that are harder to observe in ICU (eg. mobility/falls risk)
Better patient flow: allows the ICU to free a bed and avoids using ICU capacity for patients waiting for low-acuity ward care	Institution-centred bias: a decision based on bed availability is contaminated by the nonclinical pressure, i.e. it has nothing to do with the patient's interests.
Fewer care transitions: avoids another handover to a ward team which may add little value for a patient needing only prescriptions, follow-up, and discharge information.	Less mature infrastructure: ICU teams may have zero idea of what a discharge is supposed to look like, and could omit vital prescriptions or appointments.
Avoids hospital complications: less ward time may reduce exposure to delirium-provoking environments, hospital-acquired infection, falls, VTE, sleep disruption, deconditioning, MROs, etc.	Misses hospital complications: a period of hospital stay could be needed to detect complications of critical illness (DVTs, psedoaneurysms etc) which may go unrecognised in the community.
Patient and family satisfaction: the patients and their families want to go home, so naturally they are satisfied by this outcome (eg. in Lam et al, 2020)	Physician disssatisfaction with the risk: in the same study, a large percentage of physicians were uncomfortable with the idea of direct home discharge. Hilariously, most were uncomfortable because they thought the families would complain.
Potentially faster functional recovery: home may be a better environment for sleep, mobility, nutrition, family help, and return to normal roles in a genuinely recovered low-acuity patient.	Potentially slower functional recovery: Sure, you will recover faster at home if you have housing, transport, health literacy, social supports, carers, phone access, money to see a physiotherapist. Or you have none of those things, and continue to deteriorate.
Resource stewardship: may reduce unnecessary hospital days, which have a nontrivial social and personal cost; to say nothing of the opportunity cost for patients who need the ICU bed.	Resource burden export: all of the ongoing care and surveillance work is transferred to the community health providers, who have a lot less resources than hospitals,  but a lot more patients and certainly a lot less critical care experience to troubleshoot post-ICU problems.

Controversies in discharge directly home from the ICU

• The effect on patient safety is not well defined by the literature: all we can say is that it seems to be safe in Australia-like countries, provided you select patients well, and have a good community health system.
• In some health services the rate is significant (~20% in Martin et al, 2020 - a Canadian study), without significant increases in readmission rate or morbidity, i.e. it appears to be safe in well-developed health systems and for carefully selected patients
• Methodology of case selection is not well established through prospective data (but in retrospective data where success was observed, the patients were young, low acuity, and with single organ problems). 
• Whether a high rate ofd direc-home discharge should be viewed as an ICU or hospital performance indicator, is not established, but anecdotally some people already seem to view it as such; or at least, most reasonable people would agree to the counterfactual that a higher direct-home-discharge rate should not be viewed as safer, more efficient, or otherwise somehow more desirable than ward discharge.
• The quality of the data currently and historically collected makes it difficult to determine the trajectory of the rate of this.

Why might an ICU have a high rate of direct to home discharge?

Again in Question 6 from the second paper of 2025, the candidates were asked to demonstrate "a broad and high-level grasp of the more global issues" in a 80-100 words-or-less answer to this question. Fortunately, the problem of bed block is so widespread that most ICU trainees would have had zero problems creating entire sociological treatises on this topic. 

An ICU might have:

• Younger/more elective case mix (eg. mostly elective surgical hospital, a trauma bypass ED, etc)
• Local discharge criteria and ward capabilities vs. community capabilities
• Local community health support network
• Low, or high, socioeconomic status of the serviced population, which has has different implications. For example rich people are more healthy in general and require shorter ICU stay for mostly non-emergency problems and can be discharged earlier into better-resourced communities, but also it may be more lucrative to keep them in for longer to maximise the extraction of cash from their health insurance. Conversely, poor people may need to stay for longer because their access to post-ICU care will be limited and they need to be very "tidy" before they can be safely sent home, but at the same time they tend to turn up to understaffed public hospitals which are under greater bed pressure and more tempted to turf them out faster.
• Cultural factors, eg. whether the family structures locally are more, or less, likely to have extended families with more in-home care and support
• Definitions and denominators may differ, eg. jurisdictions may differ in whether they regard a transfer to a rehab facility nursing home or palliative care suite as "discharge home".
• System-level problems, eg. the availability of beds and the pressure on the ICU bed capacity can inform these decisions (though some might argue that it should never)