Regionalisation of critical care: can we sustain an intensive care unit in every hospital?

Critical care is a complex and resource-intensive discipline, that underpins many hospital services but is itself dependent on diagnostic and interventional support from other specialities. In the UK, the traditional assumption is that a hospital of any size offering acute or complex care will require, and support, an adult intensive care unit (ICU). However, this is increasingly challenged by a number of healthcare trends, including the need to sustain a specialist workforce 1 while meeting evolving critical care standards 2, 3, and the impact of reconfiguration in the many specialities that refer patients to, and support, intensive care. It may be timely to examine the drivers, potential impact and public acceptability of regionalised (‘hub and spoke’) models in adult critical care. At its simplest, regionalisation is already well established in the case of highly-specialised referral settings such as neurosciences and cardiothoracic surgery, and the specialist critical care that supports them. With such specialised services, the need for specialist rotas, and the logistical impracticality of delivering such care at every local hospital, are historically established and intuitive for both professionals and public. In a study of patients' family attitudes in one such setting, published in this issue of Anaesthesia, Chieregato and colleagues describe an attitude survey of the next of kin of 213 consecutive patients admitted to a networked neurosurgical ‘hub’ ICU in Italy 4. The service covers approximately one million inhabitants, receiving referrals from six ‘spoke’ hospitals, and transfers patients back to their originating ICU once specialist treatment is complete. This is similar to the operational function and catchment area of many UK specialist tertiary referral units. Of the next of kin of those patients returned to ‘spoke’ hospital ICUs, surveyed one year after admission, 67.4% would have preferred their family member to have remained at the specialist centre until ICU discharge, most commonly citing continuity and quality of care. The results may have been confounded in this case by the availability of certain pastoral care elements preferentially at the hub site, such as a liberal visiting policy and an emphasis on daily communication. The study does not compare next of kin characteristics, which may influence responses (for example, geographic proximity, demographics and socio-economic status). Another possible limitation is that the named next of kin may not be the most representative family member or carer. The applicability of the findings to a British setting may be subject to societal differences, with historically very high levels of expectation and loyalty toward local district general hospital care among the UK public. The authors comment that regular interaction between sites, and assurance of shared quality and continuity across a region, may help public acceptance of care pathways spanning multiple sites, pointing to clinical networks as a strategy for achieving shared standards. They also usefully highlight the often-overlooked element of post-critical care rehabilitation, now recognised as an essential element of the critical care pathway 5, which requires integration with healthcare systems close to the patient's home – a potential benefit of early repatriation. Chieregato et al.'s study examines just one aspect of the hub-spoke organisation of critical care, in the context of the highly specialised, relatively low-volume clinical setting of neurosciences, where the concentration of critical care to a few sites is driven by regionalisation of the referral speciality, and where the rationale is easily understood. More broadly, international academic discussion has focused on the case for, or against, regionalisation of critical care (the creation of hub and spoke ICUs) as a primary goal in itself, independent of referral speciality 6, 7. The evidence for clinical volume-outcome benefits in critical care is conflicting 8 and may be confounded by international variation in hospitals' organisation and staffing models. A recent UK retrospective cohort study showed a positive volume-outcome relationship in 104 844 admissions requiring mechanical ventilation in general critical care units 9. Volume-outcome benefits may be influenced by diagnostic subgroup and severity 9, 10, indicating a possible rationale for selective escalation. Conversely, a study of 5131 non-surgical admissions to US Veterans Administration hospitals, with shared organisational and governance attributes, showed unifomity of outcomes independent of case volume, suggesting that evidence-based practice and standardisation of staffing care may mitigate variation by unit size 11. Disadvantages of critical care regionalisation may include: the distance of transfer and delays in access; the risk of overwhelming the capacity of destination hospitals; strain on patients' families and carers; and the clinical risks inherent in inter-site transfer of critically ill patients, for which data are still poor in adults 12. The clinical risks may be controlled through well-organised ICU co-ordination and retrieval systems, already seen in paediatric critical care practice 13, but adult case volumes are higher and patterns of demand more complex. In practice, there are potential barriers to implementation 14, and a scarcity of evidence that reconfiguration of intensive care services improves outcomes in isolation. Primary regionalisation of adult general critical care is therefore not straightforward, and only one of a number of potential reorganisation strategies in critical care 15. We believe that changes to the landscape of adult critical care in the UK are more likely to occur as a secondary phenomenon, driven by increasing regionalisation of the clinical disciplines that refer to and support ICU. The earliest, and simplest, case of specialist tertiary services – neurosciences – has already been cited. We are now in the middle of a second wave of reconfiguration, in which services that were previously regarded as local and widely-distributed are undergoing the same transformation into networked regionalised systems. Notable examples include trauma care, stroke care, coronary angioplasty, interventional radiology and vascular surgery. This change is driven by increasing recognition of volume-outcome benefits in the primary specialities, as well as the need for specialist infrastructure and staffing, with increasing emphasis on fully sustainable subspeciality consultant rotas (rather than isolated local experts), and amplified (in the UK) by the increasingly selective placement of speciality-based postgraduate medical trainees into designated centres. Such changes impact only selectively on critical care, insofar as adequate capacity must be planned and provided at the destination hospitals and seamless referral and transfer pathways must be established to ensure that ICU patients have access to such services, regardless of site. However, in general, in this second wave of consolidation the function and viability of the referring local hospitals have not been affected, since these subspeciality patients tend to form a minority of a general hospital ICU's casemix. In the next phase, by contrast, we are likely to see more radical changes. Emerging clinical standards in areas such as unscheduled and high-risk surgery 16, 17, and changing social and political expectations toward seven-day, consultant-delivered care 18, will tend to drive increasing regionalisation of services previously regarded as core, ‘bread and butter’ clinical activity for a local general hospital. A key example is emergency abdominal surgery, which in turn has implications for acute and emergency medicine, and associated diagnostics and interventions such as 24-h imaging and endoscopy. The withdrawal of such former core specialities from some sites is likely to impact on the viability of local ICUs, both through falling caseload – with accompanying difficulties in maintaining staff census, professional skills and training status – and through the withdrawal of local clinical services that previously supported ICU patients. An emerging requirement, therefore, is to provide safe, sustainable critical care to support clinical services such as acute medicine, and elective peri-operative care of patients with co-morbidities, at sites that can no longer support a full ICU. This may be through the evolution of treat-and-transfer systems where a local hospital retains the ability to resuscitate and transfer a patient to a regional critical care centre via established local facilities and a planned pathway, or through tiered regionalisation of ICUs, such that critical care in some form can be initiated and maintained at many sites, but selected patient cohorts are transferred to high-volume, high-capability regional hubs according to risk, severity, or disease type 15. There are international precedents, most notably in healthcare systems featuring a mixture of urban centres and widely dispersed populations serviced by hub and spoke models, such as Canada and Australia, but also in the densely urbanised setting of Japan, where governmental policy, more than geography, has driven the majority of critical care to be delivered in academic centres since 1993 19. A complementary strategy may be to reorganise the ‘front door’ – in other words, to reconfigure and reduce the number of acute hospitals with emergency departments and realign them to a smaller number of high-acuity sites featuring multiple co-located, regionalised specialities, thus reducing inter-hospital transfers by admitting emergency patients directly to hospitals with the appropriate facilities. Clearly, this is a process with major financial, social and political impact, and is unlikely to be driven by critical care alone. An example is emerging in North West London, where an ambitious reconfiguration plan includes concentrating unscheduled care, currently spread across nine hospitals with emergency departments covering a population of nearly two million, into five ‘major acute’ sites and a constellation of local hospitals, elective centres, and specialist hospitals 20, with an inevitable realignment of critical care. Across the country, there is likely to be a diversity of solutions according to local circumstances and policy, varying from primary care-led local hospitals with no acute services or critical care, through to intermediate solutions involving hospitals continuing to provide selective acute and elective services, backed by well-planned access to critical care skills and resources. A particularly important question is whether acute medicine can function without a co-located ICU. Amidst this complexity, there are some simple fixed points. While generally accepting of concentration of specialist services at major hubs, the public still expects some form of local hospital care with reasonable geographic access. Whether or not every site can sustain a full ICU, it remains the case that patients with critical illness, either at admission or arising during a hospital stay, have a right to early recognition of their severity, to prompt access to critical care skills and decision-making, including appropriate admission selection, and to organised, timely access to safe facilities. Where inter-site transfers are required, they should take place via a standardised, audited system with trained staff, and destination hospitals should have sufficient capacity to assure seamless transfer by pre-agreed pathways into a designated bed, rather than ad-hoc bed-finding and delay. During their ICU stay, patients should receive best-practice care, in clinical areas that are fully compliant with extant standards and workforce requirements, and during their recovery they should benefit from an integrated approach to post-critical illness rehabilitation, working closely with services close to their home. Throughout the pathway, which may involve several sites, patients should benefit from effective handover and shared standards between sites and organisations. The solutions and means by which these goals are delivered will vary. Some may be familiar, such as the existing close professional collaboration between critical care and anaesthesia; others may involve non-traditional working patterns, such as rotation or outreach of intensivists between hub sites and spoke hospitals, including those without an ICU. Meanwhile, critical care telemedicine, while still controversial, is an increasingly well-studied modality with evidence of benefit 21, and may gain a wider mandate with new service configurations. Even currently radical options such as robotic telepresence may prove to be effective 22 and well-accepted 23. Most importantly, however, human factors, communications and culture will be essential to the success of organisational or technological changes. As suggested by Chieregato and colleagues and demonstrated elsewhere, regional networking and collaboration provides a mechanism for shared service improvement and quality assurance 24, potentially levelling differences in outcome 10. This is consistent with the successful experience of critical care networks in many parts of England and Wales since 2001. Collaboration, mutual familiarity, joint learning and a shared culture will be key to ensuring successful and seamless regional delivery of critical care, regardless of the exact configuration of facilities. Changes in clinical practice, technology, workforce and standards are driving a secular trend for medical and surgical services to become regionalised: in other words, concentrated at regional hub sites. This, in parallel with emerging standards and workforce factors, is likely to drive a trend toward regionalisation of critical care as a secondary phenomenon, side-stepping the longstanding but somewhat abstract debate over the desirability of a hub and spoke model for critical care as a primary goal. Until now, the specialist referral services driving these changes have tended to be selective in casemix and low-volume in nature, and thus have had little effect on the viability of ICU in non-hub hospitals. However it is likely that further trends toward regionalisation of high-volume, high-impact clinical activities, such as emergency surgery, will trigger more extensive changes in the critical care landscape than has formerly been seen, and will challenge current assumptions about critical care facilities and working practices for both intensivists and anaesthetists. There is a balance to be struck between access to local services, and the need for sustainable, high-quality, specialist-led healthcare that is – by its nature, and the associated workforce and logistical factors – likely to be deliverable at fewer sites than at present. The public are the ultimate stakeholders in this process – including patients' families, as Chieregato et al.'s study reminds us. Nearly one and a half decades after a UK governmental mandate to implement ‘critical care without [departmental] walls’ 25, it may be time to examine further ‘critical care across hospital perimeters’. No external funding or commercial interests declared. GS and JH hold remunerated lead roles, and AW is a salaried officer, of the North West London Critical Care Network, which is an interested party in service reconfiguration in its region. This article is written in a personal capacity, and is not intended to represent the views of the Critical Care Network or its partner organisations.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.