Jacobs Journal of Anesthesiology and Research

A Single Site Survey of Mechanical Ventilation Weaning Practice

*Diane Dennis
Department Of Physiotherapy, Sir Charles Gairdner Hospital, Western Australia, United Kingdom

*Corresponding Author:
Diane Dennis
Department Of Physiotherapy, Sir Charles Gairdner Hospital, Western Australia, United Kingdom

Published on: 2018-03-02


Introduction: Mechanical ventilation (MV) is a routine part of ICU patient management. Ventilatory support is quickly withdrawn when no longer required, however in approximately one third of patients this process is complex and prolonged. Weaning difficult patients from MV is controversial, with a number of valid approaches and little high-quality evidence. The reason for the lack of conformity in approach may be due to uncertainty regarding the best method, variable individual experience and different understanding of the application of support. Aim: The aim of this study was to describe MV weaning practice and define differences between professionals where they existed at a single site. Method: A 30 question survey was disseminated to medical, nursing and physiotherapy staff at the intensive care unit of Sir Charles Gairdner Hospital. Questions related to clinical experience, perception of current practise, and clinical rationale underlying decisions to commence and progress weaning in four patient scenarios. Results: A 48% response rate demonstrated consistency in broad weaning principles, but variation in detail. Readiness to commence weaning was indicated by the resolution of primary disease and restoration of consciousness; complex cardiac patients were most difficult to manage. The most common pattern of weaning was a gradual stepwise reduction in support on a spontaneously breathing ventilator mode such as pressure support and this occurred preferentially during daytime, with rest overnight. Most practitioners agreed that practice was variable and could be improved. Conclusion: This study described MV weaning practices at one site and may provide the basis for the development of specific MV weaning guidelines to enhance future uniformity.


Mechanical Ventilation; Ventilator Weaning; Intensive Care


Mechanical ventilation (MV) for acute respiratory failure is now a routine part of intensive care (ICU) patient management. A 2009 survey of intensive care practitioners found that 58% of patients in Australian and New Zealand ICUs received MV and it may be applied utilizing any of a number of different ventilators via numerous ventilation modes. Ventilatory support is usually withdrawn from patients when it is no longer required. In the majority of patients this is a simple matter of disconnection from the ventilator followed by observation that patient breathing is adequate and safe. This is particularly the case in patients ventilated for only a short period (hours), as for surgical procedures. This simple discontinuation of ventilation is usually undertaken either in post anaesthetic care units (recovery rooms) or in the ICU. In other patients a more complex and prolonged process of liberation from support is required, and this is referred to as “weaning” from MV. In 2005 an international task force categorized weaning as simple, difficult or prolonged. Once the underlying process necessitating MV has started to resolve, weaning should be considered, as increased duration of MV leads to a progressive rise in complications such as ventilator acquired pneumonia, sepsis and mortality. Approximately 69% of ventilated patients will be successfully weaned from support and extubated on a first attempt without difficulty, leaving around a third of ventilated patients who will present some difficulty.

Weaning difficult patients from MV is a controversial topic, with a number of valid approaches to weaning, and little high-quality evidence to support any one practice over another. Techniques used include steady reduction of mandatory respiratory rate utilising intermittent MV, the use of progressive reduction in pressure support (PS) or continuous positive airway pressure (CPAP), T piece trials and spontaneous breathing trials (SBTs). However, the application of these techniques appears to be based more on physician experience and preference rather than on robust evidence [5]. Anecdotally, practices associated with the weaning of MV vary between healthcare providers in our hospital, and approaches may change considerably from day to day. Further, many patients who are spontaneously breathing but still require some support are transferred from the ICU to step down wards – with invasive (BiPAP synchrony machine (Respironics, Pennsylvania, USA) and non-invasive ventilation (NIV) managed locally, possibly with different strategies and underlying philosophies. A recent study of our ICU cohort reported that almost one third of patients ventilated for >7 days, and surviving ICU and hospital stay were discharged from the ICU to the ward with continuing MV [6]. The reason for the lack of conformity in approach may be due to the fact that there is uncertainty about the best methods for conducting weaning [3], variable individual experience in different diseases and facilities, and different understandings of the application of MV. Whatever the reason, there is potential for continuity of care to be compromised and for this to impact on optimal and timely progression of weaning. Recognition of the common problem of variable management of prolonged MV and difficult weaning in the United States and Europe has led to the development of dedicated weaning facilities [7-12], however currently no such facilities exist in Australia. The aim of this study was to describe practices utilised in difficult weaning from MV in ICU at a single site, and to identify any differences in these practices or in the priorities of the personnel involved.

Material and Methods

This was an observational descriptive study of professionals regularly involved in the weaning of patients from MV. The study was performed with the approval of the Sir Charles Gairdner Group Human Research Ethics Committee (Study number 2011-031).


All ICU intensivists, senior registrars, senior physiotherapists and seven representative clinical nurses involved in the weaning of patients from MV were identified from departmental managers at Sir Charles Gairdner Hospital in Western Australia. Information sheets relating to the purpose and scope of the survey and questionnaires were disseminated to these people electronically. There was no written consent, as a returned completed survey and/or an agreement to interview inferred consent.

Outcome Measure

Authors developed a 30 question survey around the process of weaning that included both technical and non-technical questions. Non-technical data were collected in order to establish that those surveyed had experience and expertise in the process of weaning from MV. These data included baseline demographics, work history and employment information as well as their perception of the uniformity, effectiveness, and evidence–base of weaning practice at the study site. Technical questions were around the respondent’s individual approach to weaning patients described in four clinical scenarios. These scenarios were: 1. A 23 year old female admitted to ICU requiring mechanical ventilation via endotracheal tube post elective surgery. 2. A 30 year old female ventilated for the past 8 weeks via tracheostomy with Guillain Barre syndrome who has now progressed to recovery of hand and foot movement and some (2/5) limb girdle strength. The patient is now ventilated with Pressure Support (PS)24cmH2 O (providing a tidal volume of 6mL/Kg and respiratory rate 22/min), Continuous positive airway pressure (CPAP) 5cmH2 O, and FiO2 0.3 3. A 65 year old female mechanically ventilated after an exacerbation of chronic obstructive pulmonary disease, who has been a lifelong smoker with a baseline forced expiratory volume in one second of 1.2L. The patient has now been ventilated for 5 days via an endotracheal tube and can breathe spontaneously on PS 18cmH2 O, CPAP 10cmH2 O and FiO2 0.3 and 4. A 60 year old female who is post-myocardial infarction with significant left ventricular dysfunction. She was initially intubated and ventilated for pulmonary oedema (PO) but has been extubated and is now receiving non-invasive ventilation (NIV) via a face mask (IPAP 14cmH2 O, EPAP 8cmH2 O), is well oxygenated on 6L/min O2 and haemodynamically stable. Each time she has a trial of removal of NIV she rapidly develops PO and requires placement back onto NIV within 20 minutes. Participants rated the most important factors (1= strongly agree; 2=agree; 3= unsure; 4= disagree; 5=strongly disagree) that they would take into account in the decision to commence and progress weaning from MV.

On receipt of completed surveys, data was entered into an electronic database and basic statistical analysis were undertaken in order to summarize the data descriptively for the purposes of reporting frequencies of each response (MS excel 2003, Redmond, Washington, USA). Small numbers precluded further statistical comparison between groups.


Participant flow and characteristics

Surveys were disseminated to 23 individuals and completed responses were received from 7 medical staff, 3 physiotherapists and one nurse, reflecting a 48% overall response rate. All respondents reported experience with the Evita XL ventilator (Drägar Medical, Lübeck, Germany) and some (n=7, 64%) had experience with the BiPAP synchrony ventilator (Respironics, Pennsylvania, USA) ventilators. Five reported experience with other ventilators. Experience with weaning varied across professions but was no less than 3 years (median, range) – intensivist (25, 7-33); senior registrar (8, 3-12); physiotherapist (9, 7-25); nursing (9, 0). Five participants reported weaning experience outside Australia (UK (3), Canada (2), New Zealand (1). One participant reported past experience with a formalised written weaning protocol.

Current hospital practice

Most respondents reported a good personal understanding of MV weaning (n=9, 82%), and that hospital practice was effective (n=7, 64%). The majority of respondents believed that hospital practice of MV weaning was not uniform (n=9, 82%), could be made more uniform (n=8, 73%), and could be improved (n=8, 73%). Seven respondents (n=7, 64%) indicated that it may be possible to establish a written protocol, including the only participant with past experience in this. Most responses in favour of development of a protocol also stated that it should be more of a guideline than a strict protocol, as strict rules were unlikely to apply to all patients, or to be adhered to by all staff.

Commencement of Weaning

Those surveyed were asked what factors they would take into consideration in deciding when to commence weaning from MV in each of the four clinical scenarios presented. Responses varied between scenarios, but there seemed to be general agreement between practitioners. Factors considered important (rated 1 or 2) in a young uncomplicated patient such as post elective surgery (Table 1) were conscious state (n=9, 82%), observed resolution of the primary disease, gas exchange, haemodynamic stability (all n=8, 73%), ventilation pattern and airway evaluation (both n=7, 64%). Factors considered least important (rated 4 or 5) were evaluation of respiratory muscle strength (n=11, 100%), sleep deprivation, nutritional status and renal function (all n=10, 91%).

The most important considerations for a patient with neurologic disease (Table 2) were respiratory muscle strength, ventilation pattern (n=10, 91%) and observed resolution of the primary disease (n=9, 82%), followed by conscious state (n=8, 73%), respiratory capacity and amount and thickness of secretions (n=7, 64%). Least important (rated 4 or 5) was renal function (n=10, 91%). The most important considerations for a patient with chronic respiratory disease (Table 3) were the amount and thickness of secretions (n=9, 82%), and observed resolution of the primary disease, the ventilation pattern and conscious state (all n=8, 73%). The least important (rated 4 or 5) was nutrition and renal function (n=6, 54%). In a patient with chronic cardiac disease (Table 4), the most important considerations were the observed resolution of the primary disease (n=11, 100%), chest X-ray evaluation, fluid balance status (n=10, both 91%), ventilation pattern and time of day (both n=9, 82%). Other important considerations were haemodynamic stability, conscious state (both n=8, 73%), gas exchange, ventilator settings and staff availability (all n=7, 64%). The least important (rated 4 or 5) consideration was the psychological status (n=7, 64%) and presence of ongoing sepsis, renal function, caloric depletion and respiratory muscle strength (all n=6, 54%).

Weaning progression

There was complete agreement that in all MV weaning scenarios, progress should be reviewed face to face at least daily (n=11, 100%) and more than half (n=6, 54%) of respondents preferred review of patients more than once daily. The majority of respondents did not agree (rating 4 or 5) that MV weaning should be a slow continuous process (neurological patient n=7, 64%, chronic respiratory patient n=8, 73% and chronic cardiac patient n=9, 82%). It was preferred that weaning be an on/off process whereby strength is built up over time with periods of increased work and complete rests between these periods. There was also consensus that respiratory muscle strengthening was not the primary goal of rehabilitation (n=9, 82%). The survey asked respondents to provide a description of how they would progress weaning from MV in scenarios 2-4. No data were collected around the first scenario as it was thought this progression was more straightforward and predictable. Across scenarios 2-4, the “optimisation of underlying factors” was a common theme. Specific parameters to be optimised around each clinical scenario are summarized in Table 5. Rest from weaning overnight was also considered to be important, with return to baseline PS or mandatory MV settings, and to normalise arterial CO2 and pH. Respiratory rates of 25-30breaths/minute were targeted in each scenario. In patients with chronic respiratory disease, the use of tracheostomy in weaning was considered by more than half of the respondents (54%), although the timing of tracheosto 

                                               Table 1. Frequency of respondent’s rating for outcomes considered in the commencement of weaning in the young uncomplicated

                                               Table 2. Frequency of respondent’s rating for outcomes considered in the commencement of weaning in the patient with neurological disease

                                               Table 3. Frequency of respondent’s rating for outcomes considered in the commencement of weaning in the patient with chronic respiratory disease

                                               Table 4. Frequency of respondent’s rating for outcomes considered in the commencement of weaning in the patient with chronic cardiac condition

                                               Table 5. Progressing weaning in each clinical scenario

my was variable. Respondents indicated that tracheostomy should be performed immediately (18%), after there had been a period of failed weaning (18%), or the timing was not specified (18%). Tracheostomy was considered less in the chronic cardiac patient (18%) – one respondent indicating they would perform tracheostomy immediately (9%) and one after a week of weaning (9%).

Weaning failure

Markers of weaning failure are shown in Table 6. All respondents agreed that an altered breathing pattern indicated weaning failure (100%). Other indicators were increasing oxygen requirements (82%), asynchrony with ventilator (73%) and deterioration in ABGs, patient exhaustion or long periods of sleeping, diaphoresis and tachypnoea (all 64%).

                                                Table 6. Frequency of factors considered by respondents to indicate weaning failure.


Broad consistency in optimising the patient

Results demonstrated remarkable consistency in broad principles and approaches to weaning from MV amongst staff surveyed. There was a focus on optimising factors contributing to delays in weaning, then a fairly uniform use of decreasing levels of support in a spontaneously breathing ventilator mode. Variations in practice could be identified, but these occurred predominantly at a “detail” level in terms of approaches to ventilation and settings used and also appearing to correlate with appropriate consideration of the clinical scenarios presented. Anecdotally many of the clinicians described a hands-on approach to the titration of weaning that was difficult to reproduce by answering a questionnaire. Nonetheless, some medical staff were able to provide very specific parameters whilst others gave only very broad conceptual overviews. Across professional groups, the physiotherapists provided the most specific guidelines, however small numbers preclude direct comparisons between groups. A recently published survey of Australian and New Zealand intensivists [13] reported that measurements of respiratory function – respiratory rate, required levels of pressure support and PEEP and level of consciousness – were commonly used as predictors of both readiness for weaning and weaning success. Our survey results centred more on the broad concept of “optimising” the individual patient medically, physically, psychologically, pharmaceutically and nutritionally. The most common marker to indicate weaning readiness across all patient scenarios was the resolution of primary precipitating disease and conscious state. In addition to the broad notion of “optimising” patients prior to weaning, results here suggest different priorities may exist in optimising patients according to differing diagnostic groups. Respondents identified and mostly agreed upon clinical indicators for weaning readiness in each clinical scenario, with fewer indicators for the chronic respiratory and neurological patients compared with the chronic cardiac patient. Respiratory and neurological patient markers mostly related to the primary system affected – respiratory or neurological. Optimising a cardiac patient was considered much more complex – with more factors taken into consideration, possibly indicative of the fragility and complexity of this patient cohort. In particular, diastolic dysfunction may often be underestimated and loss or reduction of PEEP can be poorly tolerated, with this effect often not being anticipated. Co-existing conditions and deconditioning may further compromise efforts to wean from ventilatory support. 

Pattern of weaning

Respondents to the questionnaire mostly reported the use of PS ventilation in weaning from MV. The most commonly cited method of weaning from support was utilisation of spontaneous breathing on PS with either a gradual stepwise reduction in support as tolerated, or with periods of reduction in support separated by periods of rest. These findings are in keeping with previous surveys of Australian and New Zealand Intensive Care practices [2,13], where the most commonly used ventilator modes were SIMV and PS, with weaning mainly occurring by progressive reduction in support on a spontaneously breathing mode.

An additional modality discussed was the use of invasive ventilation delivered via a tracheostomy utilising a BiPAP synchrony machine (Respironics, Pennsylvania, USA) in combination with short spontaneous breathing trials (SBT). The use of SBTs is common in Australian and New Zealand ICU practice – with reportedly 35 % of ventilated patients undergoing SBTs of variable duration [2]. A recent retrospective review [6] at this facility demonstrated that 30 % of patients ventilated for more than 7 days were transferred to high dependency units from the ICU on BiPAP synchrony. After discharge from the ICU patients were then managed by respiratory physicians who also managed their weaning from MV rather than intensive care staff. Criteria described for these discharges of ventilated patients to a lower acuity setting are that patients are able to tolerate SBT for finite periods. To our knowledge, this practise has not yet been discussed in the literature, however it may facilitate ICU discharge and so cost saving, although the impact on hospital LOS, mortality and morbidity has yet to be determined.

Timing of weaning: day/night

Most respondents preferred to wean by day using an interval training model and rest at night, rather than undertake a slow continuous process. It may be that the vulnerability of the ICU patient cohort, combined with the potential for less optimal positioning, lower staffing levels and skill-mix overnight may have been taken into account. It is possible that respondents working on the wards may have reported different clinical practise, given that patients discharged from ICU will generally have achieved a certain level of robustness and endurance such that the impact of fatigue may be of less importance.

Potential for weaning guidelines

This group of highly experienced weaning practitioners were unable to provide a completely uniform approach to weaning from MV. A potential explanation is that the weaning of complex patients is an individualised process, whereby superficially similar scenarios may have considerable practical differences and require substantially differing approaches. Straightforward weaning may be a predictable and uniform practice, but more complex weaning decisions must also rely on local practices and the individual clinician’s experience and patient responses. Although highly detailed, the questionnaire utilised in this study was not interactive, and therefore could not take into account the visual and “hands-on” nature of the process of weaning which is, in itself, an evolving process. In this context only broad recommendations can be provided from the current study, and formal standardised weaning protocols for difficult weaners may be very difficult to develop and gain consensus.

Survey responder burden

The survey tool was complex so as to provide more specific evidence than that already available in the literature [14.]. Although generating interesting data, it was onerous and possibly resulted in significant responder burden which may have in part influenced the response rate. In hindsight, utilising a questionnaire may not have been the most appropriate tool to measure the “hands-on” approach that is required to titrate weaning. It may be that the use of a responsive flow diagram may have facilitated a better response rate and better universal trends and guidelines.


One of the limitations of the study is the fact that results are generated from only 11 people, and these represented a 48 % response rate whereby results may have limited generalizability. Our respondents were an eclectic group of practising clinicians across different profession groups, with perhaps less motivation to complete the survey than that of other cohorts who may have been involved in a clinical trials group [14]. Weaning at this centre is a shared responsibility, and a significant part of the intent of this survey was to describe differences between professionals where they existed. Another limitation is the unequal numbers in professional groups that made comparison difficult. This is particularly evident in nursing, where there may be no generalizability from one response. Although results largely reflect a medical/physiotherapy perspective, it should be noted that those surveyed were all senior practising staff such that findings should reflect actual ICU weaning practice at this site.

Future Trials

There was overall agreement that complex cardiac patients are the most difficult to wean. It follows that if we could establish weaning guidelines for this cohort, other cohorts may be able to fit these guidelines, albeit with faster progression. It may also be prudent to further examine our site’s practise of early discharge to step-down units both in terms of progression (ventilator settings, oxygenation targets) and outcome (time for weaning, whether or not patients are ultimately able to be completely weaned, mortality rates and readmissions to both ICU and the hospital itself).


This study adds to the limited evidence available for the prescription of weaning MV. Although numbers are small, the detail provided by the ICU-based cohort reflects current practice at a single site – highlighting similarities and differences between different professionals within the ICU. This study has allowed professionals to openly exchange their ideas and philosophies in management of patients who are difficult to wean, and may provide the basis for the development of specific weaning guidelines to enhance uniformity in the future. The survey tool used requires substantial refinement in order to collect meaningful data without significant responder burden.


The authors gratefully acknowledge all survey participants.


1. Bersten A. Mechanical Ventilation. In: Bersten A, Soni N, eds. Oh›s Intensive Care Manual. 6 ed. Oxford: Butterworth Heinemann. 366-355 ,1997.

2. Rose L, Presneill JJ, Johnston L, Nelson S, Cade JF. Ventilation and weaning practices in Australia and New Zealand. Anaesthesia And Intensive Care. 107-99 :(1)37 ,2009.

3. Boles JM, Bion J, Connors A, Herridge M, Marsh B et al. Weaning from mechanical ventilation. The European Respiratory Journal: Official Journal Of The European Society For Clinical Respiratory Physiology. 1056-1033 :(5)29 ,2007.

4. Burns KE, Adhikari N, Keenan SP, Meade M. Noninvasive positive pressure ventilation as a weaning strategy for intubated adults with respiratory failure. Cochrane Database of Systematic Reviews. 8)4 ,2010).

5. Blackwood B, Alderdice F, Burns K, Cardwell C, Lavery G, et.al. Use of weaning protocols for reducing duration of mechanical ventilation in critically ill adult patients: Cochrane systematic review and meta-analysis. BMJ ,2011 342.

6. Patman SM, Dennis, D., Hill, K. The incidence of falls in intensive care survivors. Aust Crit Care. -167 :(3)24 ,2011 174.

7. Kahn J. The evolving role of dedicated weaning facilities in critical care. Intensive Care Medicine. 10-8 :(1)36 ,2010.

8. Stoller JK, Xu M, Mascha E, Rice R. Long-term outcomes for patients discharged from a long-term hospital-based weaning unit. Chest. 1899-1892 :(5)124 ,2003.

9. Scheinhorn DJ1, Hassenpflug MS, Votto JJ, Chao DC, Epstein SK et al. Post-ICU mechanical ventilation at 23 long-term care hospitals: a multicenter outcomes study. Chest. ,2007 93-85 :(1)131.

10. Scheinhorn DJ, Hassenpflug MS, Votto JJ, et al. Ventilatordependent survivors of catastrophic illness transferred to 23 long-term care hospitals for weaning from prolonged mechanical ventilation. Chest. 84-76:(1)131;2007.

11. Iregui M, Malen J, Tuteur P, Lynch J, Holtzman M, et.al. Determinants of outcome for patients admitted to a longterm ventilator unit. Southern medical journal. :(3)95 ,2002 317-310.

12. Schönhofer B1, Euteneuer S, Nava S, Suchi S, Köhler D. Survival of mechanically ventilated patients admitted to a specialised weaning centre. Intensive Care Medicine. ,2002 916-908 :(7)28.

13. Ambrosino N, Gabbrielli L. The difficult-to-wean patient. Expert review of respiratory medicine. 692-685 :(5)4 ,2010.

14. Rose L, Blackwood B, Burns SM, Frazier SK, Egerod I. International Perspectives on the Influence of Structure and Process of Weaning From Mechanical Ventilation. American Journal of Critical Care. 1)20 ,2011): e-10e18.