Weaning units: lessons from North America?

Anaesthesia, 2007, 62, pages 374–380 doi:10.1111/j.1365-2044.2007.04984.x ..................................................................................................................................................................................................................... SPECIAL ARTICLE Weaning units: lessons from North America? C. P. Subbe,1 G. J. Criner2 and S. V. Baudouin3 1 Specialist Registrar, Thoracic & General Medicine, Advanced Trainee Intensive Care Medicine, Wrexham Maelor Hospital, Wrexham LL13 4TD, UK 2 Director, Division of Pulmonology, and Critical Care Medicine Director, Medical Intensive Care Unit and Ventilator Rehabilitation Unit, Temple University Hospital, 3401 North Broad Street, Philadelphia, PA, USA 3 Senior Lecturer in Anaesthesia and Critical Care Medicine, Department of Anaesthesia, Royal Victoria Infirmary, Queen Victoria Road, Newcastle upon Tyne NE1 4LP, UK Summary In the United Kingdom over 5% of critical care beds are occupied by stable patients weaning from mechanical ventilation. In North America, diagnosis related groups (DRGs) were introduced over a decade ago. These provided an economic impetus to develop more cost effective regional weaning centres. The imminent introduction of Payment By Results may encourage similar developments in the UK. The evidence for weaning centres is reviewed and detailed organisational and outcome data from two North American centres presented. These units differ from UK critical care units in terms of nurse : patient ratios and types and numbers of ancillary staff. Limited data, mostly from North America, suggest that weaning centres may be better at improving outcome in ventilator-dependent patients compared with standard critical care. The existing evidence is not conclusive and highlights the need for UK-based studies on organisational approaches to the provision of weaning and longer term critical care. . ...................................................................................................... Correspondence to: S. V. Baudouin E-mail: s.v.baudouin@ncl.ac.uk Accepted: 17 December 2006 The number of patients with complex medical conditions that are admitted for intensive care is increasing throughout the industrialised world [1]. In addition, the number of elderly patients with slower recoveries following an acute illness is increasing. Prolonged weaning is more likely in an elderly population [2], especially one with multiple comorbidities, and has become a major issue in the provision of critical care in Europe, as well as North America. Prolonged weaning account for 41% of total ventilator time (longer in patients with pre-existing lung disease) in one Spanish study [3]. In the UK, 6% of intensive care unit (ICU) beds in one region were occupied by patients experiencing difficulties in weaning [4]. Prolonged weaning is associated with patient-related, financial and service provision problems [5, 6]. Furthermore, the provision of intensive care beds for postoperative care of emergency or elective surgical patients, in a system with a limited and fixed number of intensive care beds, is an ongoing problem in the UK. Surgery might 374 have to be delayed at times when beds are taken up by patients with difficulty in weaning. In North America the financial problems associated with the care of long-term ventilated patients were exacerbated by the introduction of diagnosis related groups (DRGs) [7]. Re-imbursement to hospitals became predominantly dependent on the diagnosis of the patient rather than the service given or the number of procedures performed. Hospitals received single payments based on the diagnosis for long-term ventilated patients. This caused deficits in budgets of ICUs [8, 9]. This was still true after a revision of the DRG system with new codes for long-term ventilated patients with tracheostomy [10]. In the UK, the National Casemix Office has been working since 1993 on an analogous system of health care resource groups (HRGs) for re-imbursement of services in the acute sector. HRGs will form the basis for a new system of ‘payment by result’ that has been phased in since April 2003 [11], initially for specific surgical procedures,  2007 The Authors Journal compilation  2007 The Association of Anaesthetists of Great Britain and Ireland Anaesthesia, 2007, 62, pages 374–380 C. P. Subbe et al. Weaning units . .................................................................................................................................................................................................................... Æ and subsequently in other specialties. Critical care units have been asked to collect a minimum critical care dataset that replaces augmented care periods starting from April 2006 to be able to introduce HRGs for critical care from April 2008 onwards (there is still some uncertainty about the start date). National tariffs will be informed by the data collected from April 2006. Data for HRGs have already been integrated into the data set collected by the majority of British Intensive Care Units for the Intensive Care National Research and Audit Centre (ICNARC). Case-related funding could cause significant problems for patients with prolonged stay. These would become ‘outliers’ and require separate funding streams. Outliers have so far been defined for other specialties as patients that stay more then three times the average length of stay [12]. In this context it seems timely to examine models from other countries where inpatient weaning units were created under the pressure to increase cost efficiency. Financial pressures in North America helped to establish specialised Ventilator Dependent Rehabilitation Units with a lower complement of nursing staff to patients [13]. These units admit longer-stay patients (usually more than 21 days) who have failed to wean from mechanical ventilation [14]. Most patients suffer from single organ failure (respiratory) and are haemodynamically stable. The majority of reports from dedicated weaning centres for patients with prolonged ventilation are from North America. Methods One author visited two weaning units in North America with considerable academic track records: the Ventilator Rehabilitation Unit (VRU) at Temple University Hospital, Philadelphia, Pennsylvania, and the Ventilator Dependent Rehabilitation Unit (VDRU) at St Mary’s Hospital, Mayo Clinic, Rochester, Minnesota. Both of these units are located in a section of an acute care hospital that is geographically distinct from the ICU as well as the general wards. Results Observations from two US weaning units Overview of care delivery There are significant operational differences between UK and North American Critical Care units. Whereas the majority of ICUs in the UK operate at a staffing level of one nurse per patient, American units often have one nurse for every two patients supported by a team of respiratory therapists that take on all aspects of care related to set-up and running of the ventilation. The number of  2007 The Authors Journal compilation  2007 The Association of Anaesthetists of Great Britain and Ireland Critical Care beds per patient population is also significantly higher in North America. Whereas most UK ICUs are ‘closed’ in the sense that the intensive care physicians take a lead in the care of all patients, in the USA there is a co-existence of ‘closed’ units with ‘open’ units, where a general physician or surgeon will be responsible for the care of their own patients on the unit. In the North American weaning units the majority of patients have tracheostomies or are undergoing postextubation rehabilitation. Respiratory therapists undertake a large part of the care related to mechanical ventilation that in the UK is performed by nurses. These are physiotherapists with 2 years of training, who set up and maintain the ventilators and perform chest physiotherapy. At any given time including nights and weekends there are up to two respiratory therapists on the unit. In addition a number of physical therapists are attached to the units. Physiotherapists who deal with the nonrespiratory part of the therapy are called physical therapists in the North America. Physical therapists train for 7 years and are in charge of the exercise program for muscle strength, balance and co-ordination. Case mix Some 40% of the patients at the Temple University VRU and 66% at Mayo VDRU are postsurgical, commonly following thoracic and cardio-thoracic surgery [16]. Increasingly, the patients have morbid obesity or lung pathology including chronic obstructive pulmonary disease (COPD) and pulmonary fibrosis. Admission criteria for the VRU include the following: patients should have failed two weaning attempts on an ICU, are haemodynamically stable, on minimal or no inotropic drugs and alert enough to be able to participate in rehabilitation. Patients should be able to breathe for a short period unaided in case of disconnection of the tracheostomy tube. Patients with multi-organ failure or malignancy with limited prognosis are usually not admitted. VRU, Temple University Hospital, Philadelphia The VRU at Temple University Hospital, Philadelphia, is an 18-bedded unit founded in 1988. Patients are referred from ICUs within Temple University Hospital (70% of admissions to the VRU) and from surrounding hospitals across several states. Costs for the majority of patients are re-imbursed by insurance. Patients are accommodated in single- or two-bed rooms. Every patient has a clock opposite his bed next to a detailed timetable for the day. Routine monitoring of patients consists of pulse oximetry complemented by telemetric ECG monitoring as required. Some patients have peripheral central venous lines for antibiotics. 375 C. P. Subbe et al. Weaning units Anaesthesia, 2007, 62, pages 374–380 . .................................................................................................................................................................................................................... Æ Table 1 Staffing levels (WTE) at Temple University VRU. Setting Temple University VRU: non-acute cases only No. of beds Staffing day shift Nurses Physiotherapists Medical staff Clerical staff Total staffing Nursing staff** Physiotherapy** 18 Medical staff Clerical staff Mayo clinic VDRU 9 4 + 1 Pulmonary nurse specialist during standard working hours. 4–5 registered nurses during standard working hours. 5 (2 respiratory therapists + 3 physical therapists) 3 (2 respiratory therapists + 1 physical therapists). 2 (1 senior medical staff + 1 junior medical staff) 2 (1 senior medical staff + 1 junior medical staff) 2 1 28 15† 2 (1 senior medical staff + 1 junior medical staff)* 2 28 Physiotherapy staff are part of the large critical care units and are deployed flexibly depending on work load 2 (1 senior medical staff + 1 junior medical staff)* 1 Staffing levels are adjusted for UK employment practices. *Assuming night time cover by medical staff from other services. **Assuming standard contracts and 7 WTE to staff one bed taking into account 168 h per week cover plus cover for leave and absence. †24-h staffing for respiratory therapists and office hours staffing for physical therapists and UK style physiotherapists. WTE, whole time equivalent Central venous pressure and intra-arterial pressure are not routinely measured. Monitoring of respiratory rate is performed as part of bedside observations. Respiratory therapists or residents take arterial blood samples as required. Venous samples are taken by a phlebotomy service. Staffing The unit is staffed by registered nurses at a ratio of one nurse for four patients (Table 1). There are two therapists that work exclusively on the unit. An assistant or physiotherapy student supports them on most days. Medical staff consists of a Fellow (comparable to a Specialist Registrar) and the supervising Attending (comparable to a UK consultant) who does daily rounds. All Attendings of the VRU at Temple are pulmonologists and critical care physicians. A Pulmonary nurse specialist (comparable to a UK nurse practitioner) co-ordinates admissions, discharges and performs some of the unit specific procedures, such as tracheostomy change, during standard working hours. Sessional commitments by a speech therapist, occupational therapist, pharmacist, social worker, clinical psychologist and case managers complete the team. The multidisciplinary team meets once a week with private health care and equipment providers to discuss all patients, training progress and plan discharges. Weaning methods The treatment of patients admitted to the VRU consists of a detailed medical workup and intensive respiratory and physical rehabilitation. Pharmacological therapy is similar to UK practice, but continuous nebulisation [15] is more frequently performed. Rehabilitation consists of respiratory and physical rehabilitation. 376 Respiratory rehabilitation Respiratory rehabilitation is performed by respiratory therapists. The main mode of weaning at Temple University Hospital is volume-controlled ventilation with one to two daily T-piece trials of increasing length. Non-invasive ventilation is used as a supplementary means of weaning in patients following removal of tracheostomy tubes. Standard ICU ventilators are supplemented by ventilators used exclusively for noninvasive ventilation. Removal of secretions is aided by physiotherapy delivered by hand and with a vibrating pneumatic jacket (Theravest, The Vest Hill Rom, St Paul, MN, USA). Prior to decannulation the tracheostomy is often downsized to a small, non-cuffed and capped tube exclusively for aspiration of secretions (comparable to the usage of Minitrach (Portex, Hythe, Kent)). Patients are encouraged to perform exercises to strengthen inspiratory muscles by breathing for 10–15 min at a time through a valve with adjustable resistance (Threshold IMT, Respironics Inc., NJ, USA). Muscle strength is initially assessed by measuring maximum inspiratory pressure with a manometer. The patient is then encouraged to perform exercises at 30–40% of maximum inspiratory pressure. While patients have a tracheostomy, a Passy-Muir valve, or an electro-larynx and ⁄ or buccal resonator is offered to improve verbal communication. General rehabilitation General rehabilitation starts with a feeding programme that might take into account measurements of resting energy expenditure. Long-term ventilated patients are often fed through percutaneous feeding tubes (gastric or jejunal). Once respiratory function progresses, the swal 2007 The Authors Journal compilation  2007 The Association of Anaesthetists of Great Britain and Ireland Anaesthesia, 2007, 62, pages 374–380 C. P. Subbe et al. Weaning units . .................................................................................................................................................................................................................... Æ low is assessed fluoroscopically and patients will then be encouraged to eat even if the tracheostomy is still in situ. Ventilated and non-ventilated patients are aggressively mobilised by physical therapists. They are mobilised as early as possible and often on the day of their admission. If patients are able to stand they will be encouraged to walk with a Zimmer frame or stick while being ventilated with a transportable ventilator. Once or twice a day patients are being transferred to the gym of the VRU. Training includes cycling exercises for arms and legs. On average the VRU admits one to two new patients per week for weaning purposes. Beds are used flexibly and other patients who are not weaning may be admitted. In the case of Temple University these are patients with lung– and heart–lung transplants and patients following lung-volume reduction surgery. Empty beds can be used for other medical ‘outliers’ but are used predominantly for patients with respiratory pathology, i.e. asthmatics, patients on long-term non-invasive ventilation, continuous positive airway pressure (CPAP) or long-term ventilation admitted for other reasons. Demographics During the period from January 2000 to June 2004 the VRU ventilated 232 patients. Mean length of hospital stay was 35 days (standard deviation (SD) 36 days) Mean length of stay in the VRU was 21 days (SD 24 days). Three-year survival of patients discharged from hospital was 60% (personal communication). Ventilator Dependent Rehabilitation Unit, Mayo Clinic, Rochester, Minnesota The VDRU is located in the vicinity of a large general ICU. The VDRU is a closed unit with the patients being under the care of the lead physician. The VDRU has an impressive academic track record. Due to the detailed data collection over a long period of time it has been used as a model for weaning units across the US and for cost analysis. In the first 5 years the VDRU admitted 206 newly ventilator-dependent patients, 190 of whom survived to discharge. Seventy-seven percent were able to return to their homes and only 37 were either completely or partially ventilator dependent at discharge. The 4-year survival was 53% [17]. The VDRU is organised in a similar way to the VRU at Temple University. The unit has nine beds that are used for a mix of weaning and pulmonary High Dependency ⁄ Step Down patients. The Mayo Clinic patients are somewhat unusual in that a high percentage of patients are out-of-area medical referrals or selfreferred patients. Invasive ventilation is provided for patients with tracheostomies and non-invasive ventilation for patients following exacerbations of COPD.  2007 The Authors Journal compilation  2007 The Association of Anaesthetists of Great Britain and Ireland Staffing levels are higher than at Temple University, with one nurse per two patients. The level of physiotherapy input is high but the physiotherapists are shared with other departments. The medical staffing consists of an attending and a pulmonary fellow. In both VDRU and VRU, teaching is an integral part of the daily rounds. For weaning, the Mayo unit uses collapsible tracheostomy tubes (Bivona Medical Technologies, Gary, IN) that if deflated allow the patient to breathe and speak against minimal resistance. Tracheostomies are capped prior to decannulation and oxygen is entrained through an inlet in the cap. A tilt table is used for initial mobilisation of patients who have been bed bound. This aims to improve the orthostatic circulatory response until patients are able to reach 90-degree positions without a significant drop in blood pressure. Evidence review There is a limited evidence base to support the service model of a weaning unit as opposed to standard intensive care. There is no published randomised controlled trial of patients with prolonged ventilation comparing care in a dedicated weaning centre with standard care. The published literature is almost exclusively from North American centres [17–21] and comparability is therefore difficult because of differences in intensive care case mix and financial incentives for certain models of care. However, financial incentives might become more comparable with the current introduction of payment by results in the UK. A demonstration prepared for the US Office of Research and Demonstrations Health Care Financing Administration in 1996 [22] compared prospectively admitted patients in four established weaning units with data from a large research database of unselected Medicare patients admitted and ventilated for more than 21 days. Interpretation of the results is difficult due to the heterogeneity of the main outcomes reported. Two of the four weaning units (Temple University and Mayo Clinic) appeared to significantly out-perform standard ICU care in terms of patient outcome and survival (Fig. 1). However, survival in the other two units was comparable to the control population. Three of the four weaning units also achieved considerably better home discharge rates compared to standard care (Fig. 2). However, median length of stay in an acute care facility was longer in these patients, resulting in increased costs per patient episode (86 days in all weaning facilities as compared with 52 days in the control sample). Cost per day in the participating weaning units was estimated at between US$958 and $2064 per day. At the time this was $500 to $700 cheaper than an intensive care bed for comparable patients. These estimates did not take into account the total cost of care. Costs were not offset against savings in 377 C. P. Subbe et al. Weaning units Anaesthesia, 2007, 62, pages 374–380 . .................................................................................................................................................................................................................... Æ 600 Ventilator days Hospital stay Survival Days 400 200 0 Control data Unit 1 Unit 2 Unit 3 Unit 4 The North American study found increased lengths of hospital stay in patients who successfully weaned (Fig. 1). Unfortunately, an improved quality of care may not be cost-neutral for patients with long-term respiratory needs. There is only limited research that has examined whether the additional cost translates into a physical and emotional health that is acceptable for patients and funding bodies. The health status of survivors seems to be comparable with that of other patients groups with chronic disease with reasonable emotional wellbeing and limitations in physical activities [23, 24]. Figure 1 Outcome data from the four North American Weaning units which took part in the study on weaning units. The four intervention groups are patients from four weaning units that were funded for the purpose of the study to wean patients with more then 21 days of ventilation. Median length of survival from admission, median hospital stay and median duration of ventilator episode in days are presented. The data are derived from the demonstration undertaken by the Office of Research and Demonstrations Health Care Financing Administration. The control group was chosen from a large research database of US intensive care patients. Two of the four units (Units 1 and 4) performed significantly better when compared with non-specialist units. Percentage of patients discharged home 100 75 50 25 0 Control data Unit 1 Unit 2 Unit 3 Unit 4 Figure 2 Percentage of patients discharged to their own home, either self-caring or cared for by family. Data are derived from the demonstration undertaken by the Office of Research and Demonstrations Health Care Financing Administration. long-term facilities. Although the hospital stay for patients treated in a weaning facility was longer, the greater independence at discharge might be able to offset cost in the long run in a system such as the National Health System (NHS) where hospital and nursing home care are largely funded centrally. The effect of a weaning unit on availability of critical care beds was also not part of the demonstration. It is therefore not possible to compare the absolute cost of the different care models. This makes it difficult to estimate total costs and savings in the NHS for similar care. It should also be acknowledged that the total costs in a health care system with weaning units might increase. 378 Discussion Application of the weaning unit model to the UK? A review of data from the National Intensive Care case mix program (ICNARC, personal communication) looking at 278 012 admissions to 170 critical care units in the UK from December 1995 to January 2004 found that 10 579 ventilated patients (3.8% of all admissions) were still in the ICU 21 days following admission. Information from the case mix program does not differentiate between patients still ventilated after 21 days or those no longer ventilated. These prolonged admissions had a mean age of 60.3 years (SD 16.2 years) and 61% were male. The most common primary reasons for admission were pneumonia (18.6%), aortic or iliac dissection or aneurysm (4.9%), septic shock (4.8%) and acute pancreatitis (2.9%). Mortality at discharge from ICU was 23.4% and at discharge from hospital 39.8%. Survivors and non-survivors stayed a median of 29 days in ICU. Hospital survivors stayed an additional 28 days in hospital and non-survivors an additional 18 days. Clearly, not all of these patients could (or should) be transferred to a North American-style weaning centre. A 12-month prospective survey in the North of England identified 161 patients who still required ventilatory support after 2 weeks of critical care, but were otherwise stable [4]. Only 20 of these patients were ventilated for more then 28 days. However, they occupied 6% of available intensive care beds in the region. To date there have been few publications on specialised weaning programs in the UK [24]. A recent study found the costs of a bed in a weaning centre to be significantly lower than in a comparable ICU bed (cost per day 1350 Euro). Thirty-eight percent of admitted patients were weaned. The mortality of these patients was comparable with that of other ICU patients [25]. Conclusions The North American Weaning Centre, as exemplified by two high-quality units visited by one of the authors,  2007 The Authors Journal compilation  2007 The Association of Anaesthetists of Great Britain and Ireland Anaesthesia, 2007, 62, pages 374–380 C. P. Subbe et al. Weaning units . .................................................................................................................................................................................................................... Æ provides one model for UK practice. Organisational change is never completely evidence-based, but several factors support the establishment of such centres in the UK. These include arguments based on better outcome and quality of care, more efficient use of critical care resources, concentration of specialist resources and expertise. However progress in the UK is slow due to several issues: • Despite significant improvements in critical care training, many units still rely on input from specialists with limited knowledge and experience in the management of longer term problems resulting from intensive care. • Despite the much publicised modernisation agenda, the NHS remains very reluctant to embrace change. A lack of a true health care market combined with an emphasis on cost control prevents innovative (and possibly cheaper) approaches to care. • The fragmentation of the NHS into overlapping Hospital Trusts, Primary care trusts (PCTs) and Strategic Health Authorities hinders the development of new Regional or Super-regional approaches to care delivery. Weaning centres aim to move patients with complex care needs successfully from ICU to the community. By definition this requires the co-ordination of several services. Current NHS organisation hampers and even prevents this process. • Change in the NHS often occurs by central government directive. Interest in Critical Care and therefore weaning has significantly diminished since the number of patients needing long-distance transfer fell substantially following increased funding. Specialised knowledge and skills resulting in specialised care for patients with difficulty in weaning from mechanical ventilation are potentially as useful for the care of patients in the UK as for the care of North American patients. However, given the absence of data from trials and the presence of centralised funding of the health service in the UK it would seem advisable to fund trials of different models of service delivery for patients on prolonged mechanical ventilation. 2 3 4 5 6 7 8 9 10 11 12 13 14 Acknowledgements Dr C. P. Subbe received a travel grant from the ScaddingMorriston-Davies Joint Fellowship in Respiratory Medicine to investigate the function and functioning of weaning centres in the USA. 15 16 References 1 Berenson RA. Intensive Care Units: Clinical Outcomes, Costs, and Decision Making (Health Technology Case Study 28).  2007 The Authors Journal compilation  2007 The Association of Anaesthetists of Great Britain and Ireland 17 Prepared for the Office of Technology Assessment. US Congress. OTA-HCS, November 1984. Washington, DC: OTA, 1984. Ely EW, Wheeler AP, Thompson BT, Ancukiewicz M, Steinberg KP, Bernard GR. Recovery rate and prognosis in older persons who develop acute lung injury and the acute respiratory distress syndrome. Annals of International Medicine 1992; 136: 25–36. Esteban A, Alia I, Ibanez J, Benito S, Tobin MJ. Modes of mechanical ventilation and weaning. A national survey of Spanish hospitals. The Spanish Lung Failure Collaborative Group. Chest 1994; 106: 1188–93. Robson V, Poynter J, Lawler PG, Baudouin SV. The need for a regional weaning centre, a one-year survey of intensive care weaning delay in the Northern Region of England. Anaesthesia 2003; 58: 161–82. Scheinhorn DJ, Chao DC, Stearn-Hassenpflug M. Liberation from prolonged mechanical ventilation. Critical Care Clinics 2002; 18: 569–95. Dasgupta A, Rice R, Mascha E, Litaker D, Stoller JK. Four-year experience with a unit for long-term ventilation (respiratory special care unit) at the Cleveland Clinic Foundation. Chest 1999; 116: 447–55. Fetter R. Background. In: Fetter R, Brand D, Gamache D, eds. DRGs: Their Design and Development. Ann Arbor: Health Administration Press, 1991. Gracey DR, Gillespie DG, Nobrega F, Naessens JM, Krishan I. Financial implications of prolonged ventilator care of medicare patients under the prospective payment system: a multicentre study. Chest 1987; 91: 424–7. Dracey DR, Nobrega FT, Naessens JM, Krishan I. Financial implications of prolonged ventilator care under DRGs 474 and 475. Chest 1989; 96: 193–4. Gracey DR. The problem with diagnosis related Group 475. Chest 2002; 122: 299–301. Healthcare Resource Groups. The Health and Social Care Information Centre. http://www.ic.nhs.uk/casemix. Implementing Payment by Result. London: Department of Health, 2004. Gracey DR, Hardy DC, Koenig GE. The chronic ventilator-dependent unit: a lower-cost alternative to intensive care. Mayo Clinical Proceedings 2000; 75: 445–9. Make BJ, Hill NS, Goldberg AI, et al. Mechanical Ventilation beyond the intensive care unit. Report of a Consensus Conference of the American College of Chest Physicians. Chest 1998; 113 (5 Suppl.): 289S–344S. Anonymous. British Guideline on the Management of Asthma. London: British Thoracic Society and Scottish Intercollegiate Guidelines Network, 2003. Gracey DR, Naessens JM, Viggiano RW, Koenig GE, Silverstein MD, Hubmeyer RD. Outcome of patients cared for in a Ventilator-Dependent Unit in a General Hospital. Chest 1995; 107: 494–9. Gracey DR, Hardy DC, Naessens JM, Silverstein MD, Hubmayr RD. The Mayo ventilator-dependent 379 C. P. Subbe et al. Weaning units Anaesthesia, 2007, 62, pages 374–380 . .................................................................................................................................................................................................................... Æ 18 19 20 21 rehabilitation unit: a 5-year experience. Mayo Clinic Proceedings 1997; 92: 13–9. Modawal A, Candadai NP, Mandell KM, et al. Weaning success among ventilator-dependent patients in a rehabilitation facility. Archives of Physical and Medical Rehabilitation 2002; 83: 154–7. Latrino B, McCauley P, Astiz ME, Greenbaum D, Rackow EC. Non ICU care of hemodynamically stable mechanically ventilated patients. Chest 1996; 109: 1591–6. Stoller JK, Xu M, Mascha E, Rice R. Long-term outcomes for patients discharged from a long-term hospital based weaning unit. Chest 2003; 124: 1892–9. Scheinhorn DJ, Artinian BM, Catlin JL. Weaning from prolonged mechanical ventilation: the experience at a regional weaning centre. Chest 1994; 105: 534–9. 380 22 Lewin-VHI, Inc. Ventilator Dependent Unit Demonstration; Outcome Evaluation and Assessment of Post Acute Care. Washington DC, USA: Office of Research and Demonstrations, Health Care Financing Administration, 1996. 23 Chatila W, Kreimer DT, Criner GJ. Quality of life in survivors of prolonged mechanical ventilator supprt. Critical Care Medicine 2001; 29: 737–42. 24 Smith IE, Shneerson JM. A progressive care programme for prolonged ventilatory failure: analysis of outcome. British Journal of Anaesthesia 1995; 75: 399–404. 25 Pilcher DV, Bailey MJ, Treacher DF, Hamid S, Williams AJ, Davidson AC. Outcomes, cost and long term survival of patients referred to a regional weaning centre. Thorax 2005; 60: 187–92.  2007 The Authors Journal compilation  2007 The Association of Anaesthetists of Great Britain and Ireland