Guidance. Adult Critical Care Strategic Development. A Framework for North West London. North West London. May 2012 (revised)

Transcription

1 North West London Adult Critical Care Strategic Development A Framework for North West London Guidance Originally authored in 2008, revised May 2011 and May 2012 May 2012 (revised)

2 Adult Critical Care Strategic Development - A Framework for North West London - Guidance - Introduction 1. This guidance provides specialist clinical input into the critical care implications of acute service reconfiguration in North West London. It is produced in response to the publication and ensuing discussions about the Healthcare for London report A Framework for Action produced by Professor Darzi in Revisions (2011) are informed by NW London regional reconfiguration discussions, in which the Network has participated. In 2012, further revisions exploring level 2 plus in detail have been added. 2. The document is authored on behalf of the North West London Critical Care Network Steering Group, with wide discussion and input from Network clinical groups, together with Commissioner and acute Trust executive representation. Purpose 3. Any detailed implementation of acute service changes will rely on decision-making at multiple levels, including internal service reconfigurations within acute Trusts, collaborative commissioning between acute Trusts and PCTs; and strategic support from London-wide agencies. 4. This document is intended to inform this decision-making process at all levels. The document does not set out to say which acute service changes should or should not occur, but instead to provide informed insight into the patterns of critical care provision which would be required to support each change. 5. Critical care reconfiguration poses challenges, as critically ill patients are intrinsically clinically unstable. Patient needs and clinical condition can alter rapidly and unpredictably, requiring a degree of contingency planning to underpin each acute care model. A key intention of this document is to ensure that the critical care impact of any acute service change is constructively and comprehensively considered. 6. The document also considers issues which are not specifically raised in A Framework for Action but where reconfiguration provides a further opportunity to improve patient outcome through improved access to local and tertiary critical care, seamless referral pathways for specialist care such as emergency neurosurgery, and safe inter-hospital transfers where necessary. 7. The success of this guidance, and of the Critical Care Network s input, relies on its readership being prepared to actively project its key points and recommendations into all relevant acute service discussions and fora. 8. The document is not intended to survey or designate services per site, nor to state the critical care capacity required to support a particular service change, since this needs more detailed consideration of each proposal. The Network would Page 2 of 20

3 expect this guidance to be considered in detail as part of the overall case and service model for specific service change proposals. Clinical scope 9. The document is already in bullet point and grid format, and further summary may risk the loss of information. The list of headings below is to provide an outline of the overall content rather than an executive summary. The services and themes discussed within the document are: Page 4 Page 7 Surgery: Major themes: theatre and critical care cover for selective surgical take and non-major surgery vs. unselected surgical take and abdominal surgery. Transfer pathways, capacity, co-location, and stepdown for specialist surgery (vascular, cardiothoracic, neurosurgery, burns). Medicine: Major themes: critical care requirements for selected vs. unselected acute medical take, impacting on A&E reconfiguration. Escalation, stabilisation facilities, and referral routes. Requirement for ring-fenced referral beds at supporting site. Critical care support for interventional cardiology, neurology, and specialist medicine. Page 10 Maternity services: matching capacity to rising demand and complexity of deliveries in London, with increased workload on local DGH units. Page 11 Specialist intensive care tertiary referral pathways for advanced respiratory support, weaning from mechanical ventilation, and renal support. Page 13 Opportunities to improve care through seamless referral pathways, spanning pre-hospital diagnosis, centralised triage, improved bed-finding and capacity. Page 13 Cross-speciality factors enabling safe and effective service provision: training (including minimum sustainable unit size), information technology and telemedicine, transfer systems, and rehabilitation. Page 16 Audit, governance, and quality assessment. Page 17 Glossary Explanation: Levels of Critical Care : L2 = HDU or High Dependency Unit / Care L2+ = Standalone L2 with short-term L3 escalation capacity and close support from a paired L3 unit L3 = ITU or Intensive Care Unit Page 18 Appendix 1: Level 2+ - an emerging model Page 3 of 20

4 Activity Key clinical issues Supporting measures required to enable facility reduction (where applicable) Recommendations 1. Surgery 1.1a Elective, nonmajor surgery. (e.g.: day surgery, laparoscopic cholecystectomy). Medically fit (ASA 1/2) patients undergoing this category of surgery will not require planned L2 or L3 care. However: Patients may unexpectedly need critical care following surgery. Immediate complications are likely to happen in safe environment (theatres/ recovery) with skilled staff and monitoring. Delayed complications may occur out of hours in ward environment and will be especially life-threatening in absence of L3 facilities. close links and shared working between critical care and Recovery. Evidence-based systems for early detection of patient deterioration. Robust call-out systems and availability of anaesthetic or critical care expertise on-site. Facilities for on-site resuscitation and stabilisation Level 2 cover may be adequate if enabling requirements satisfied 24/7 theatre availability may not be necessary, unless these are the only resuscitation/ stabilisation facilities on site. Reduction to L2 requires reciprocal investment to provide guaranteed L3 capacity at a designated referral centre. Robust audit and incident monitoring system to detect operational problems. 1.2 Major abdominal surgery (includes unselected surgical emergency take) Elective: as 1.1, but case volume and increased physiological complexity = greater likelihood of perioperative complications and unplanned L2 need. Pre-operative screening and optimisation, and planned L2 provision, may reduce unplanned L2 and L3 requirement Units undertaking major emergency and elective abdominal surgery should be carried out in presence of L3 critical care facilities. Page 4 of 20

5 Emergency: peri-operative complications that do occur are more likely to be severe and life-threatening, therefore significant expected need for L3. While L3 must be provided, demand can be effectively reduced through effective preoptimisation and post-operative L2 care. Hospitals accepting unselected surgical take should be considered in this category. Peri-operative optimization facilities (L2) should be planned 24 hr theatre availability required. 1.3 Specialist surgery vascular Some major elective vascular surgery needs planned L3 care. Many emergency major vascular cases need L3 care (e.g. ruptured aneurysm). n/a L3 required (but see below) Vascular needs L3, including adequate fixed or expansible capability when on-call for other sites. n/a L3 required Adequate L2 provision may spare some L3 need. Most vascular patients have comorbidity, and may have additional need for L2 or L3 care on this basis. n/a L3 required (but see below) Needs to be co-located with specialist diagnostics + interventional radiology. In some cases, planned preoptimisation and elective post-op L2 care can minimize LOS and need for higher levels of care. Pre-operative assessment, optimisation, and elective post-op L2 reduced L3 requirement 1.4 Specialist surgery - cardiothoracic Elective and emergency requires L3 n/a specialist L3 required optimise L3 capacity via step-down, rehab, repatriation pathways Page 5 of 20

6 Some patients arrive unstable (worse outcomes, worse postop LOS)?due to lack of intraaortic balloon pump (IABP) at referring sites Ambulance compatibility with IABP and increased use at referring sites, may improve tertiary L3 capacity by decreasing LOS at destination Review IABP provision and usage at outlying sites; and ambulance compatibility (specialist critical care transfer vehicles) 1.5 Specialist surgery - neurosurgery (elective and emergency) Limited capacity can hinder fulfillment of clinical access targets (e.g. 4 hours from diagnosis to treatment of a neurosurgical emergency) Delayed referral transfer may increase post-op length of stay and worsen outcome Critical are capacity may be determined by operational factors as well as funding: e.g. ability to step down to specialist L2 beds or return to base hospital. Provision of adequate L3 capacity, including demand surges Shared care pathway with referring sites for prompt diagnosis, stabilisation, transfer Specialist L2 facilities, and shared care pathways (incl. rehabilitation/ repatriation) with referring hospitals, will maximise availability of specialist L3 capacity. Advantage of dedicated stepdown/rehab pathway: tertiary centre will be discharging to known unit with lower risk of deterioration or bounceback. Concentration of expertise. Advantage of repatriation to referring hospital: integration with local community care. Specialist L3 facilities required (neurosurgical ITU) Specialist step-down (L2) required Either robust cross-trust repatriation policies, or sufficient local step-down capacity Needs to be co-located with specialist diagnostics + interventional radiology. Robust IT and other links may reduce overall length of stay through speedier referral & intervention < 4 hrs 1.6 Specialist surgery - burns Burns patients can be ventilated in general units but there are significant outcome gains from early surgical intervention Adequate ITU provision within designated burns centres so that requirement for L3 care does not cause delays in transfer and surgery Excellent pre-transfer care at referring centres Capacity planning and shared protocols for early referral and transfer Robust protocols for early management of burns at referring hospitals Page 6 of 20

7 Activity Key clinical issues Supporting measures required to enable facility reduction (where applicable) Recommendations 2. Medicine 2.1 Unselected acute general medical take Includes GI haemorrhage, myocardial infarct, acute CVA, abdominal pain. These may require specialist backup (surgery, angiography) and the decision to continue unselected take should be reexamined if these are lacking. Safety patients in high dependency who then deteriorate and require mechanical ventilation, may be transferred at time of highest risk and instability. Capacity Many patients can be managed at level 2 when there is safe L3 escalation cover on-site. These may instead tend to be committed to pre-emptive ventilation and transfer when there is no L3 facility on site. Gatekeeping particularly identification of cases where intensive care may be futile. This is a difficult aspect of critical care, needing training, experience, and time to discuss with family and referring teams If this is lacking, risk that larger numbers of untreatable patients than at present will be committed to intensive care. Sites without L3 facilities would need to be staffed with emergency specialists with critical care expertise. Recruitment to such posts and retention of skills in a non-l3 site is problematic. Availability of expertise/ reassurance (?telemedicine/rotation/medical outreach from L3 sites/other) Withdrawal/refusal decisions are harder to make at a distance than active management decisions. Trained intensivist presence required on-site. Active end-of-life pathway provision at ward level may partially mitigate the risk of increased futile admissions from non-l3 units. It is recommended that all units continuing to accept unselected medical take should retain full L3 facilities as a default, but a detailed, site-specific risk assessment may show that Level 2+ (L2 plus capacity to escalate to L3 for hours) is safe and feasible. This will depend on relatively low volume and on all factors addressed in section 2.2 being provided. Sites undertaking selective medical take (see 2.2) should retain Level 2+ critical care as a minimum. Decision to continue unselected medical take (and consequences for critical care cover) should be taken in light of co-located diagnostics and support services. Page 7 of 20

8 2.2 Selective acute general medical take e.g. sites excluding GI haemorrhage or abdo pain; or sites which do not admit chest pain or CVA. Reasons for selecting/restricting medical takes, might include lack of backup (surgical service, or interventional cardiac or neuro-angiography). Safety, capacity, and gatekeeping concerns as in 2.1 still apply However, selective medical take will mean lower likelihood and volume of patients requiring escalation to L3, altering the risk profile. Selective medical take is likely to be associated with lack of out of hours surgery A hospital running primarily with L2 facilities would still need to retain capacity to escalate individual patients to L3 for limited periods. Where there is restricted or no out of hours surgery, alternative escalation and stabilisation facilities may be under-used and available (theatres, anaesthetic rooms, anaesthetists) but such cover should be used on a planned, protocolised, funded basis. Anaesthetic depts may form an alternative tier of consultants to supervise gatekeeping decisions for short-term escalations with a view to next morning transfer. Training, intensivist backup from L3 sites, and rotations to maintain experience may be required. L3 patients will need a destination to transfer to. Unlike the norm in London, this should be a planned, designated pathway to an identified hospital (not determined by random bed availability) which has enough capacity to guarantee a bed to the referring hospital when needed. This model works well in cancer hospitals in Manchester and London (Royal Marsden: Sutton and Fulham sites). Similarly, the destination unit must be able to repatriate referrals seamlessly back to the originating hospital for step-down once safe to do so. Doing so to a known unit, with a designated return bed and close communication links to the L3 centre, reduces the risk of secondary deterioration. An L2+ unit may be suitable with selective medical admissions and very infrequent L3 need. A L2+ unit provides L2 care as a routine, but retains the ability to escalate a single pt to L3 and stabilise for up to 24h pending transfer, in a suitable L2+/3 bay (not ad hoc, theatres, etc.) Escalation route should be clearly defined i.e. Hospital A (L2+) always refers to Hospital B (L3), with strong comms, staff rotation, and training links. Generic EBS referral when an L3 bed is needed, is inadequate: there should be guaranteed ring-fenced capacity at Hospital B for its paired L2+ unit. Reducing to L2+ may be unsafe if expected transfer volume is high. This should be carefully modelled. Volume of activity should be taken into consideration: a system that works well with intermitted escalation/transfer, may be unsafe if the system is excessively loaded. Page 8 of 20

9 2.3 Interventional cardiology Patients with acute myocardial infarcts will be increasingly concentrated at specialist sites providing interventional angiography. A proportion of these patients will suffer ventricular failure and cardiogenic shock, or complications of angiography Angio sites may be co-located with cardiothoracic surgery (see 1.4) or may be standalone L3 facilities may be required urgently, in patients who may be too unstable to transfer L2-plus-escalation may be inadequate in this context, given severity and timescale of deterioration Intra-aortic balloon pump facilities are becoming part of the accepted standard of care for ICUs accepting patients in cardiogenic shock due to coronary disease Hospitals admitting acute MI patients should have L3 facilities. ICUs at such sites should be able to provide intra-aortic balloon pump support and access to angiography, or defined arrangements for rapid transfer to a suitable site 2.4 Neurological intervention CVA management will increasingly feature interventional strategies cf. cardiology (e.g. cerebral angiography). CVA management currently supportive/rehab based. Demand on critical care low; patients often not considered for admission as no benefit from adding critical care to current treatment Hospitals admitting acute CVA patients should develop L3 facilities alongside new interventional strategies. Emerging supportive strategies may require critical care (e.g. elective hypothermia) L3 facilities may be required under new methodologies; may need to be concentrated at acute specialise treatment centres as for Specialist medicine e.g. haematology/ oncology, immunology/ rheumatology/ specialist respiratory Patients undergoing specialist treatment (e.g. chemo or immune therapy) may deteriorate rapidly from infection or severe side effects of some class of drug. Such patients may be a challenge for ITUs without inhouse expertise Other patients may be at intrinsically higher risk of needing crit care (e.g. Specialist medical units will tend to be in large hospitals with full L3. Some units may be in standalone cancer centres or other facilities away from main sites. Such units may most appropriately operate L2 facilities, but need L3 escalation pathways similar to those discussed for selective medical take. Concentration of high-risk chemotherapy and other medical interventions at sites with specific expertise, including specialist L3 units, may be Specialities with high-risk treatments best clustered at tertiary sites with L3. Otherwise, clear resuscitation/ escalation pathways to be established, cf. item 2.2. i.e. L3 escalation facilities on site + designated destination unit for transfer + adequate capacity in Page 9 of 20

10 3. Maternity services 3.1 Consultant-led obstetric units respiratory disorders) an advantage system All active obstetric units will have a baseline rate of serious morbidity requiring critical care, e.g. from major haemorrhage or embolism. As units consolidate and become larger, there is a trend for activity and case complexity to be repatriated from specialist centres to large DGH units. There is no future for standalone consultant-led units all will be on acute hospital sites. Proximity to neonatology, acute medicine and surgery, and other hospital specialities is required for all full, consultant-led obstetric units. Critical care capacity will go along with this, but specific extra provision will need to be made for L2 and L3 workload arising from obstetrics Post-partum mothers requiring critical care should not be transferred between hospitals due to lack of capacity, both for safety and compassionate reasons (separation from child in Neonatal unit). Critical care capacity must keep pace with changes in obstetric activity. This may include specialist L2 (obstetric HDU/step-down) and/or general L2/3 (HDU/ICU) facilities in the main hospital. Local ICU capacity should be adequate to assure L3 access on demand, without interhospital transfer. 3.1 Alongside midwife-led units Selected low-risk cases. Will share acute hospital facilities of the above. Full functional adjacency (not just shared site) should ensure full access for resuscitation and critical care delivery by acute hospital team. As above 3.2 Standalone midwife-led units These units represent an obsolete model and should no longer exist. Risks include delays in transfer to sites offering appropriate levels of critical care, obstetrics, or neonatology. Standalone low-risk units must be linked on a blue-light basis to an acute hospital with full obstetric and neonatal facilities not a local non-obstetric hospital, even if nearer. Robust protocols for comms and facilities for hyper-urgent transfer to appropriate, designated site. The emergency referral route of each standalone unit must be reassessed if levels of critical care, obstetrics, or neonatology are altered in neighbouring acute hospitals. Page 10 of 20

11 4. Specialist intensive care (ITU-to-ITU referral) 4.1 Acute specialist respiratory Acute lung injury patients [needs rubric for non-clinicians] benefit from protocolised care at local units. Complex acute lung injury cases may benefit from specialist ITU-to-ITU referral. Pulmonary hypertension, adult congenital heart disease, and Lung injury management is now standard at most ITUs but may not achieve excellence at all sites. There is scope for further standardisation of referral criteria, and an agreed referral destination for West London. These patients are complex to transfer safely Local management protocols +/- outreach from specialist centre Agreed referral criteria for specialist centre Funded capacity at specialist centre to support W London needs Access to specialist critical care transfer vehicles (cf. intra-aortic balloon pumps in 1.4) 4.2 Respiratory weaning Patients requiring long-term ventilation may benefit from a specialist progressive care programme (ref. Comp.Critical Care). Such a programme may be delivered in local ITUs or specialist units (e.g. RBH, GSTT). Current capacity exists but is focussed on specialist conditions (e.g. neuromuscular disease) and not common pathology (severe COPD). Referral pathways should be to designated nearby centres. This maximises communication links and familiarity between units, enables step-down support, and minimises travel for carers. These patients may have complex domiciliary care needs. Specialist units, not being the patient s local hospital, may suffer from lack of agreed rehab and social care pathways with local PCTs which can dramatically delay discharge to home. Provision of long-term ventilation and progressive care programme with adequate capacity for W London. Standardised agreements for funding of long term domiciliary care Such transfer also frees acute general ITU beds for incoming acute pts. Page 11 of 20

12 4.3 Renal support Every general ITU has a number of referrals of ward patients needing single-organ renal support. Some patients currently admitted to ICU present as L3, but this may be avoidable if access is improved (ward patients treated promptly for acute single-organ renal failure may avoid escalation to multiple organ failure). Significant numbers of L3 patients recover to the stage of needing renal support only, but continue to take up full L3 beds. The renal unit at Imperial lacks capacity to deal with acute renal failure patients on a 24/7 basis, and accepts specialist referrals only. There may be a role for an acute renal support facility for NW London, accepting single-organ failure patients acutely from general wards around the sector. This may be run as part of a general ITU, providing ability to escalate to other organ support as required. Such a unit would operate in concert with, rather than competing with, the specialist tertiary referrral unit at Imperial, where capacity to take acute referrals may be limited by transplant and other specialist activity. Onward referral for specialist renal care would be managed by protocol. Modern ICU renal support technology may offer facility to undertake intermittent renal support in general ICUs (i.e. few hours a day rather than continuous support). Regionally: consider acute renal support facility. Requires sectorwide commissioning. Locally: acute ICUs may offer intermittent renal support with current technology, i.e. stabilise ward patients without committing them to 24/7 full ICU care. Requires local re-configuration (e.g. set aside 1 bed for renal support, with turnover of >1 pt/day). Page 12 of 20

13 5. Critical Care - First responders Roadside / Blue light call critical care Patients requiring specialist care (neuro, burns, etc.) currently taken to (usually) nearest A&E, diagnosed and then require secondary transfer for specialist surgery or ITU. This causes delay and risk to patient as well as duplication and waste of resources. Discussions between Critical Care Network and local Emergency Care clinicians indicate a move toward expert pre-hospital emergency care ( Doctor on a motorbike model) to supplement/integrate with paramedical emergency response Collaboration between Emergency and Critical Care systems will be essential to ensure that patients are triaged to the centres best equipped to deal with them e.g. multiple trauma or other multi-organ system illness. Critical Care first responder pathways should initiate early interventions and triage to most appropriate specialist service or specialist intensive care. Appropriate diagnostics and triage in the field with immediate selection of appropriate referral site, may make an important contribution to improved, seamless neuro, trauma, burns, and other pathways (see relevant sections above). 6. Notes on factors affecting configuration (across all service types) 6.1 Training requirements Critical care requires expert knowledge and training in all professions. Training requirements set an effective minimum size on sustainable ITUs. For nursing staff, access to training (e.g. ITU course) can be provided from any unit with adequate funding and backfill availability, but there is a critical mass (unquantified but estimatable at 4 L3 beds or less) at which point experience base and skill mix becomes difficult to sustain. Similar considerations apply to AHPs, although they may be better able to maintain skill mix through access to a broader range of facilities across the site (e.g. acute respiratory medicine). For medical staff consultant appointments and trainee placements are regulated by the Intercollegiate Board of Training in Intensive Care Medicine (IBTICM) which recommends that specialists work in units that see >200 level 3 admissions/year. Therefore, as well as answering the question, does this site need L3/L2 facilities?, commissioners and providers need to address the question, is the unit big enough to be sustainable?. Failure to address this may harm either recruitment, or more directly patient safety through loss of skills, or both. Page 13 of 20

14 One management strategy to address this might be rotation of staff between larger and smaller units but this needs to be carefully managed. 6.2 Information technology Discussion of reconfiguration should include consideration of telemedicine. Telemedicine which provides remote access to monitoring and diagnostics is in use in various parts of the UK (e.g. Manchester, Frimley Park) and has several benefits in more rapid access to senior opinion, more effective communication between units (e.g. for transferred patients) 6.3 Transfer mechanisms Live bed-state software (whereby each unit can see the activity and bed availability of others) has been used successfully via Charing Cross/Hammersmith (Acubase), in SW London (Ward Watcher?), and elsewhere (ICBIS in Manchester) to enable easier transfer and capacity management, and should be incorporated into ICT planning. The use of telemedicine to enable reduced medical staffing levels, as practiced in the US (e.g. VISICU) has been examined in the UK. The Network clinicians did not see a role for it when it was presented within NW London. National work, including an external review of its validity in Lothian, has replicated this. The broad consensus so far has been that such e-icu systems provide a remote expert advisory service, but this is of limited use when any intervention has to be performed by on-site staff. There may be a role for VISICU or similar systems in using ICU expertise to monitor L2 patients in other areas, to detect and prevent early physiological deterioration. One site (St. Mary s) has configured CCTV cabling and other remote monitoring facilities into its forthcoming ITU rebuild, and may be seen as a pilot site for such development within W London. Intrinsic to any new system is that some transfer activity will inevitably remain, and appropriately so (e.g. escalation to higher levels of care, large units with concentrations of expertise, or specialist units). Safety of all aspects of transfer is paramount. Existing Network workstreams on transfer safety, training, audit, and equipment will continue. The large drop in non-clinical transfers under the current system should be noted. Anticipated volume of transfers should be modelled as part of any reconfiguration that involves reducing critical care at a given site (see 2.2): the safety of any treat and transfer system will rely on its ability to cope with the activity, and the safety and cost of (particularly) high-volume transfers is a major factor in the feasibility of any reconfiguration. Routes of transfer need to be considered. Other cities in the UK use clear, designated escalation routes whereby one identified hospital always refers to another, and the second hospital always has capacity. By contrast, London currently runs on a lottery system whereby all referral units run at maximal capacity and patients go to the nearest available bed on the day. The risk profile of transfers, and local capacity decisions, are all affected by this. Page 14 of 20

15 De-escalation pathways need to be similarly clear and consistent. E.g. when oncology patients from the Christie Hospital in Manchester need level 3 they are always transferred to the Royal Infirmary, but equally they are always stepped back down to the Christie HDU after recovery, so that the Infirmary clinicians are always discharging to a known unit with full confidence and very little deterioration and re-referral. For specialist cases (e.g. neuro, burns, chest injury), considerable work has already been done in NW London, with shared protocols between referring and referral hospitals (e.g. protocols for care of brain injury patients in non-neuro hospitals). This will be extended and reinforced to ensure seamless care across a range of tertiary referral areas. On-site diagnosis and triage direct to tertiary centres by LAS and medical first responders (see item 5) would further reduce delay. It would be synergistic with a centralised retrieval team, i.e. direct tertiary triage would reduce the numbers of secondary interhospital transfers, while centralised retrieval would improve safety of those cases still needing transfer. A centralised transfer/retrieval system for adult critical care has been used successfully elsewhere in the UK and internationally. It requires pooled investment and planning between Trusts, but offers advantages of patient safety (consistency of training, equipment, and governance); better use of resources (acute sites will not intermittently lose a tier of staff on transfer); and improved outcome (shown in paediatric models). However, workload is likely to be higher for adult than for paediatric services. Transfer activity and safety will be critically dependent on the quality of local assessment and gatekeeping at the referring sites, and this should be objectively assessed (and strengthened if necessary) as part of any reconfiguration. 6.4 Rehabilitation and domestic support In addition to the specific respiratory weaning support described in 4.2, ICU and hospital bed-days may be saved through a comprehensive, cross-pct programme to offer domestic support to patients who do not recover fully from critical illness and have long-term high dependency needs. These patients are a minority but can consume a disproportionate fraction of resources: e.g. long-term occupation of ICU beds after reaching a plateau of recovery. This can be exacerbated by geographical health and social service boundaries, when the patient is in an acute hospital which is distant to their home address. Solutions may range from community support and rehab pathways, through to full technical support for non-invasive ventilation, specialist nutrition, or other needs. Discharge delays may be avoided by developing Sector- or city-wide pathways and template funding agreements so that these do not need to be negotiated for each individual case. Page 15 of 20

16 7. Notes on governance and audit 7.1 Quality assessment In 2007 a joint Network meeting of clinical and commissioning representatives agreed the principle that critical care funding (including comparative funding of each site) should be informed by a mutually agreed quality assessment framework, based on international best practice and indices. It was agreed that simple outcome measures (SMR, length of stay) were insufficient in isolation and that a structure, process, outcome outline was a likely basis for development. The Network will develop this, or equivalent, in 2008 for the 2009/10 Commissioning cycle. This approach was ratified in March 2008 at a meeting with the Collaborative Commissioning Group for NW London. The quality measures have been in place and reported since Further copies of this document may be obtained from North West London Critical Care Network Room 112 C/O NHS Ealing (PCT) 1 Armstrong Way Southall UB2 4SA +44 (0) critcarenetworknwl@nhs.net Page 16 of 20

17 Glossary: Levels of critical care and descriptions Level 0 Requires hospitalisation needs can be met through normal ward care. Level 1 a) Patients recently discharged from a higher level of care. b) Patients in need of additional monitoring, clinical input or advice. c) Patients requiring critical care outreach service support. d) Patients requiring staff with special expertise and/or additional facilities for at least one aspect of critical care delivered in a general ward environment. Level 2 ( HDU) Patients needing single organ system monitoring and support. (Patients in need of advanced respiratory support as the only major organ system supported due to an acute illness would normally satisfy the criteria for level 3). b) Patients needing pre-operative optimisation: Requiring invasive monitoring and treatment to improve organ function. c) Patients needing extended post-operative care: Extended postoperative observation is required either because of the nature of the procedure and/or the patient s condition. Included in this group would be patients needing short term, i.e. less than 24 hours, routine postoperative ventilation who are otherwise well with no other organ dysfunction, e.g. fast track cardiac surgery patients. d) Patients needing a greater degree of observation and monitoring. Observation and monitoring that cannot be safely provided at level 1 or below, judged on the basis of clinical circumstances and ward resources. e) Patients moving to step-down care. (i.e from a higher level). f) Patients with major uncorrected physiological abnormalities: These physiological abnormalities, if incorrected, are likely to indicate a patient requiring at least level 2 care. Patients with lesser degree of abnormality or other physiological abnormalities may also require level 2 or 3 care. Level 3 (ITU) a) Patients needing advanced respiratory monitoring and support: Excluded from this group would be patients needing short term, i.e. less than 24 hours, routine postoperative ventilation who are otherwise well with no other organ dysfunction, e.g. fast track cardiac surgery patients. If ventilatory support exceeds 24 hours, or other significant organ dysfunction develops, these patients now need level 3 care. b) Patients needing monitoring and support for two or more organ systems, one of which may be basic or advanced respiratory support. c) Patients with chronic impairment of one or more organ systems sufficient to restrict daily activities (co-morbidity) and who require support for an acute reversible failure of another organ system. Source: Department of Health Critical Care Information Advisory Group. January 2006 Levels of Care: (summarised from: Levels of Critical Care for Adult Patients, Intensive Care Society, 2002, see full document for further detail and clinical examples). Note that some definitions, particularly at level 1 and 2 are the subject of current review. Page 17 of 20

18 Appendix 1: Level 2+: an emerging model? 1. In a complex and evolving health economy such as NW London, with multiple acute sites and where centralization of some services may leave other sites with a more restricted case-mix, there is a potential conflict between (a) the need to provide adequate critical care to support clinical services, especially acute medicine, and (b) the consequences of keeping a full ICU running while workload shrinks to below sustainability for a safe unit (>200 L3 admissions/year). 2. A potential solution, which needs to be carefully modeled for volumes and co-dependencies per site, is that of a Level 2+ unit which allows short-term (<24h) escalation to level Classically, L2 units (HDUs) are normally complementary to full L3 units (ICUs) on the same site, with the HDU either based on acute wards to act as feeder/step-down units into a central ICU; or operating in co-location with the ICU to provide flexible and efficient options for step-up step-down care. 4. Within NW London, there are very few models of L2 care being provided in isolation, i.e. without a full ICU on site. Only one is within the NHS and is in a specialist hospital without an ED, while the other, in the independent sector, has since been upgraded to full L3. Where such units have existed, they have always had a step-up facility to L3 care including ventilation. Local models consist of: - The Royal Marsden s Sutton branch has an L2 unit which can accept and ventilate L3 patients on a short-term basis for stabilization and transfer to the main site. Staffing is provided by nursing staff transferred from the main site, plus local anaesthetics cover. - The Clementine Churchill (independent) historically operated an L2 unit with elective ventilation capacity for post-surgical patients under direct anaesthetic management. Unplanned cases needed full intensive care were stabilized and transferred to an NHS unit according to availability through the Emergency Bed Service. The Clementine Churchill Hospital has now been upgraded to full L3 capability in order to service specialist surgery and acute medical inpatient admissions. - Notably, neither hospital combined open acute medical admissions with an L2-only model: RMH does not have an A&E, and the Clementine Churchill did not accept acute medical inpatients prior to upgrading its unit. Neither hospital has a full A&E. 5. Outside London, there are models of hospitals with complex patients and unplanned ICU admissions, which operate an L2 + short-term ventilation model. - The Christie cancer hospital in Manchester operates an L2 unit with capability to escalate to L3 care for 24 hours. This enables critically ill, specialized (cancer) patients to be stabilized and transferred to an L3 bed at a designated partner site (Manchester Royal Infirmary). - Ability to ventilate in a semi-planned manner (i.e. not in theatres, anaesthetic rooms, or recovery, but in an equipped critical care facility) for up to 24 hours, ensures that patients can be fully stabilized before transfer, and that when appropriate transfer can await full (daytime) staffing levels. Page 18 of 20

19 - The small DGH at Grantham is linked to the larger Lincoln Hospital. Grantham has an open A&E, with low volumes, and is staffed with general anaesthetic cover. Lincoln has a double tier of ICU consultants (2 x 1:4); any patient requiring L3 at Grantham is retrieved by a consultant within the same shift. This is an expensive and (locally viewed as) unsustainable rota. 6. Both models differ in key respects from current patterns of referral within London - Both models rely on designated beds being available in the larger unit, and not on a free for all depending on random bed availability in any of many surrounding hospitals. The current EBS referral model in London is designed for ad hoc non-clinical transfers only. A level 2+ unit would rely on funded, ring-fenced capacity at a defined, neighbouring L3 unit. - Paired units have established referral links, and close clinical co-operation. The larger unit is able to provide on-demand critical care escalation without delays or excessive referral hurdles, using shared admissions criteria. Patients can be stepped down to the smaller unit in a planned and trusted fashion, with both units knowing each other s capabilities, assured by shared training and rotation. 7. Recommendation - An L2+ unit may be considered to support a site with one or more of o o o Elective intermediate/complex surgery which does not carry a major risk of shock/multi-organ failure Selected medical take which excludes abdominal pathology, major haemorrhage including GI, and major cardio-respiratory and neurological emergencies Specialist medicine (e.g. cancer) without an ED. - Such a unit would require, as a minimum: o o o o o o Facilities for escalation to L3, including a designated and fully equipped bay Pairing with a designated referral centre with L3 facilities, which has ring-fenced capacity for patients from the L2+ unit. Skilled staffing backup, with current experience of L3: either by rotation or on-call from main site Access to middle-grade anaesthetic cover, backed by consultant anaesthetist decision-making and access to intensivist opinion from the main L3 site Excellent communications, diagnostic, and transport links to main site Commitment to identical admission criteria to main site Page 19 of 20

20 Risks - The L2+ model is relatively untested in any hospital with ED facilities, although models are emerging (e.g. Solihull). - An L2+ model will be functional and safe only when the casemix of the hospital site means that emergency escalation to L3 is a rare clinical exception. This assumption may break down either if complexity is too high (e.g. a sick inpatient population with frequent escalation), or if volumes are higher than the model anticipates. - An L2+ unit that is in constant or regular operation at L3 level, is functioning by default as full ICU that has failed the size criteria for training, safety, and sustainability, and would require urgent review. - The array of measures required to render an L2+ model safe, may make it unaffordable. This must be weighed against the need to provide acute general services at this site, and their dependence on critical care support. Page 20 of 20