Using the prone position for ventilated patients with respiratory failure: a review

Authors


AD Wright, MRes Programme, School of Health Sciences, University of Liverpool, Liverpool L69 3GB, UK
E-mail:adwright@liverpool.ac.uk

Abstract

Aims: This review explored the evidence relating to prone positioning in ventilated patients diagnosed with respiratory failure, including acute lung injury (ALI) or adult respiratory distress syndrome (ARDS).

Background: Mortality rates for ventilated patients with ALI or ARDS are high, and there is a growing body of evidence suggesting that the position these patients are nursed in may influence clinical outcomes. However, there are no guidelines to inform nursing practice in positioning these patients.

Method: Medline, Scopus, Cinahl and the Cochrane Library databases were searched for original research reports or systematic reviews of evidence between 2000 and 2009. Reference lists of retrieved papers were hand searched and included studies were analysed using the Critical Appraisal and Skills Programme framework. A narrative data synthesis considered the strengths and limitations of studies, and findings were collated and interpreted.

Results: Application of the search strategy identified a systematic review, currently underway, which has not yet reported and 14 relevant studies eligible for inclusion in this review. Analysis showed considerable variation in study design, but suggests that PaO2/FiO2 ratio, incidence of VAP and mortality may be positively affected by prone positioning.

Conclusions: Evidence of the clinical benefits associated with prone positioning is inconclusive and provides little guidance for nursing practice. There is a need for further research into the clinical outcomes of prone positioning, and greater understanding of the practicalities of prone positioning critically ill patients is required.

Relevance to clinical practice: Nurses have a central role to play in the continual assessment and management of this patient group, including the position they are nursed in, not only to ensure the best clinical outcomes but also to provide care and comfort to the patient and their family. It is therefore important that their nursing practice and interventions are informed by the best available evidence.

INTRODUCTION

Evidence-based practice is a core theme of UK health policy (DH 2006) and presents an ongoing challenge to nurses and other health professionals (Craig and Smythe, 2007). Although practical and theoretical issues have been identified, which impact the way nurses can use evidence to support their clinical decisions and professional working (Thompson et al., 2001; McCaughan et al., 2002; Rolfe and Gardner, 2006; Meijers et al., 2006; Spenceley et al., 2007), the parameters of what constitutes best nursing evidence in the intensive care environment remain debatable. The complexity of the needs of patients in intensive care units (ICUs) and the nature of inter-professional working relationships make it difficult to distinguish which aspects of patient care should be the primary responsibility of any one professional group. However, positioning the patient in bed to prevent the development of pressure sores, maintain skin integrity, provide comfort and facilitate recovery has always been a key nursing responsibility, so it is arguable that this should be an area where the best evidence is used to inform nursing decision-making and care.

As a result of factors normally associated with their critical illness itself or the treatment they are receiving, many patients within ICU are nursed in a supine position. However, it is known that the prolonged use of the supine position increases the risk of complications such as thrombosis, atelectasis, pneumonia and the development of pressure ulcers (Shapiro and Broccard, 1997). Alternative positions for nursing ventilated patients include semirecumbent, left lateral, right lateral and prone. Periodic adjustment of a patient's position may be something that nurses do manually, but in the case of prone positioning the patient should be nursed on an appropriate pressure-relieving mattress (Rowe, 2004). Low air loss mattresses are preferable because of the potentially prolonged period in the prone position incorporating a 2–4 h repositioning regime using the swimmers position.

The utility of prone positioning in ventilated ICU patients was first explored in the 1970s. Piehl and Brown (1976) found that prone positioning of patients resulted in improvements in arterial oxygenation while Douglas et al. (1977) also suggested that prone positioning, with the upper thorax and pelvis supported and the abdomen free, resulted in an increase in the partial pressure of arterial oxygen. However, it should be noted that both these early studies had small sample sizes (n = 5 and n = 6, respectively). As is often the case in research exploring new issues, these sample sizes are not necessarily representative but did offer indications for further research.

Although the maintenance of a good PaO2 measurement is important in the care of ventilated patients, there are many other clinical benefits that are assumed to be related to prone positioning. These include increased functional residual volume, alterations in diaphragmatic movement, postural drainage of water and exudates, accelerated secretion removal, shunting of perfusion and improved ventilation (Lan and He, 2009). All these factors are important in facilitating recovery in critically ill patients, so nursing patients in the optimum position could have significant and measurable impacts on the outcome of their stay in ICU.

A particular challenge to ICU nurses is the effective care of those ventilated patients who are suffering from acute lung injury (ALI) or adult respiratory distress syndrome (ARDS). The mortality risks of these patients are particularly high, with that of ARDS reported between 35% and 45% (Phua et al., 2009). It is perhaps not surprising that more recent research exploring the effects of prone positioning in the ICU has focused on patients with these diagnoses as defined by the American-European Consensus Conference (AECC) (Bernard et al., 1994).

Although the research on prone positioning in patients who are ventilated has illustrated improvements in arterial oxygenation, the impact of prone positioning on ventilator-associated pneumonia (VAP), mortality, ICU stay and adverse events is less clear (Dodek et al., 2004; Hess, 2005; Abroug et al., 2008; Tiruvoipati et al., 2008). This lack of clarity stimulated a review of the evidence and raised questions as to which patients would be expected to benefit from being nursed in a prone position and at what point in their ICU patient career any positional changes should be instigated.

AIM OF THE REVIEW

In the absence of current guidelines for best nursing practice, the aim of this review was therefore to evaluate the evidence relating to prone positioning in ventilated patients with a diagnosis of ALI or ARDS.

METHOD

The review method incorporated a search of Medline, Scopus, Cinahl and the Cochrane Library databases, using the keywords prone* OR proning OR prone position* AND ventilat* OR artificial respirat* OR mechanical respirat* AND research OR stud* OR clinical trial*.

Studies were identified on the basis of the apparent relevance of their titles and abstracts and those that met the inclusion criteria were retrieved (Table 1). The reference lists of all retrieved papers were then examined for further sources.

Table 1.  Inclusion/exclusion criteria
InclusionExclusion
• Published between 1 January 2000 and 31 December 2009• Published before 1 January 2000 or after 31 December 2009
• Written in English• Not written in English
• Related to adult humans with ALI or ARDS• Related to animals, paediatric populations, non-ALI or non-ARDS
• Related to prone positioning in the ICU environment• Not related to the prone position and non-ICU setting
• Physiological or clinical outcome measures reported• Physiological or clinical outcome measures not reported
• Systematic reviews or research studies• Editorials, letters or discursive pieces

The retrieved research reports were evaluated using the framework of the Critical Appraisal and Skills Programme guidelines for critical appraisal (Public Health Resources Unit, 2007). The use of this tool helped provide a consistent and systematic evaluation of each paper and an indication of the methodological quality of the studies. A range of research designs were reported on the retrieved studies; therefore, no attempt was made to extract and pool data. A narrative analysis and synthesis of data were undertaken, which involves the consideration of strengths and limitations of reported studies and the collation and interpretation of findings from studies that have been included in a review (Centre for Reviews and Dissemination, 2009). Using the narrative approach, the implications of these study findings for nurses caring for ventilated patients were summarized.

RESULTS

The Cochrane Database of Reviews identified one intervention protocol in the anaesthesia review group (Bloomfield et al., 2009), which has not yet reported their results. Initial searching identified 267 published papers, which appeared relevant to the field of inquiry and application of the inclusion criteria identified 14 research reports, of variable design, which were eligible for inclusion in this review. The hand search of reference lists did not identify any additional material. Table 2 shows a summary of the included studies.

Table 2.  Studies included in the review
ReferencesDesignStudy populationInterventionMain findings
  1. ALI, acute lung injury; ARDS, adult respiratory distress syndrome; ARF, acute respiratory failure; EVLWI, extravascular lung water index; FiO2, fraction of inspired oxygen; GCS, Glasgow coma scale; LOS, length of stay; PCV, pressure controlled ventilation; PEEP, positive end expiratory pressure; PP, prone position; SP, supine position; ETT, endo tracheal tube; IAP, Intra abdominal pressure; PS, Pressure sores.

Johannigman et al. (2000)Prospective, observational study20 ARDS patientsAll PP, >6 h PP and <5 h SP90% responded with increase PaO2 on first PP; no ETT tubes, central or arterial lines displaced; 8 peripheral lines and 1 nasogastric tube displaced; PS in 5 PP patients
Gattinoni et al. (2001)Multicenter randomized control trial304 ALI PEEP >10 or ARDS PEEP >5152 SP versus 152 PP, 6 h/day 10 days PPAdverse events were similar in PP and SP; oxygenation improves without affecting mortality
Hering et al. (2001)Prospective study16 mechanically ventilated ALI within 24 h of ALI diagnosisAll PP 3 h PP and SP in random orderArterial oxygenation and systemic blood flow improves with PP; slight increase in IAP during PP
Papzian et al. (2001)Prospective cohort study49 ARDS patientsAll PP, 6 h PPPP induced a significant increase in the PaO2/FiO2 ratio p < 0·001
Venet et al. (2001)Retrospective observational study59 hypoxaemic, ventilated patients PaO2/FiO2 ratio <300 mmHg; non-ARDS and ARDS patients PaO2/FiO2 ratio <200 mmHgAll PP, >4 h PPImproved PaO2/FiO2 ratio in 90% of PP periods; PP in first 7 days of ARDS diagnosis related to decrease in mortality; adverse events not noted
L’Her et al. (2002)Prospective observational study51 intubated and mechanically ventilated ARDS patientsAll PP, 12 h/day PP96% positive response in PaO2/FiO2 ratio; accidental extubation (1), line twisted resulting in low BP (2), lines removed (3), enteral feed stopped (13) because of vomit (12%), because of increased aspiration (14%), PS 45%, 20% grade III or IV
McAuley et al. (2002)Prospective observational study11 ALI/ARDS PaO2/FiO2 ratio <300 mmHg despite 2:1 ratio PCV and PEEP >10All PP, 18 h PPEVLWI decrease significant at 18 h; improved PaO2/FiO2 ratio; facial oedema in all patients
Gattinoni et al. (2003)Retrospective analysis225 (209 patients analysed) ALI PEEP >10 or ARDS PEEP >5All PP, 6 h/day 10 days PPPatients that reacted with a decrease in CO2 demonstrated an improved survival outcome; no adverse events reported
Guerin et al. (2004)Prospective, unblinded, multicenter controlled trial791 hypoxaemic ARF adults >18 years; PaO2/FiO2 ratio >300 mmHg; mechanically ventilated >48 h413 PP versus 378 SP, >8 h/day PPPP may reduce VAP but has significantly harmful effects with no effect on mortality; PS and ETT obstruction higher in PP than SP but no figures or severity documented
Voggenreiter et al. (2005)Prospective randomized trial40 ALI >24 h and PEEP >5 or ARDS >8 h and PEEP >521 PP versus 19 SP, >8 h and <23 h PPAdverse events no significant difference; ARDS and pneumonia reduced in PP with no effect on mechanical ventilation or mortality
Mancebo et al. (2006)Multicenter randomized control trial136 intubated, ventilated ARDS patients; adults >18 years76 PP versus 60 SP, 17 h/day 10 days1 ETT tube kinked, 1 unplanned extubation, 1 Swan Ganz dislodged, 1 HF line malfunction, 1 catheter and 1 NGT accidentally removed; PP 15% absolute decrease in mortality, did not reach statistical significance because of small sample size p < 0·12
Rossetti et al. (2006)Non-randomized, open prospective controlled trial41 ARDS and FiO2 >0·5 to maintain PaO2 >80 mmHgAll PP, 3 h/day78% showed improvement in PaO2/FiO2 ratio >15% of baseline; weight significantly higher in responders (p = 0·058); accidental extubation (1), PS not noted
Davis et al. (2007)Retrospective review61 trauma and surgical patients with ALI/ARDS44SP versus 13PP, 4 crossover to PP 4 h rota PPReduced ventilator days and overall hospital LOS in PP compared with SP; mortality reduced in PP p < 0·01
Fernandez et al. (2008)Multicenter, open randomized clinical trial40 mechanically ventilated ARDS <48 h of ARDS diagnosis21 PP versus 19 SP, >20 h/day untilPaO2/FiO2 ratio higher in PP, statistically significant at day 3; 15% absolute increase in 60-day survival in PP but not significant because of small sample; accidental extubations similar in groups, PS increased in PP but no figures or grades noted
Romero et al. (2009)Pilot feasibility study15 severe ARDS PaO2/FiO2 ratio <100 mmHg; adults >18 years; ventilation <72 h; oxygenation index >15All PP, <48 hPP is safe and improves oxygenation in severe ARDS; grade II PS (2), no lines displaced or extubations

ANALYSIS

Although the volume of publications in this field of inquiry since 2000 suggests a growing interest in the position in which ventilated patients with an ALI or ARDS diagnosis should be nursed, the inclusion of only 14 studies in this review highlights some of the difficulties of clinical research in this critically ill patient population. Only six of these reported studies were trial designs, and none of these were able to recruit the number of patients necessary to demonstrate statistical power to provide robust evidence of effectiveness of the prone positioning intervention. The included studies also reported the measurement of a range of clinical outcomes, and those which may be of particular interest to ICU nurses are reported here.

PaO2/FiO2 ratio

An important clinical benefit of an improved PaO2/FiO2 ratio is a reduction in the classification of respiratory failure. This may reduce the risk of developing further complications associated with ARDS, such as multi-organ failure, which is one of the leading causes of death in this patient group. All included studies demonstrated an increase in PaO2/FiO2 ratio in the groups of patients who were nursed in the prone position; however, there was considerable variation in the way improvement was defined and the time allowed before a response to prone positioning was measured. In a prospective cohort study of 49 patients with ARDS, Papazian et al. (2001) defined ‘rapid’ responses as those showing a 20% improvement in PaO2/FiO2 ratio within 1 h of prone position and ‘slow’ responses as those achieving this within 6 h of being nursed in the prone position. Although the majority the patients (73%) in this study were categorized as rapid responders, it was interesting to note that they also demonstrated a significant reduction in the PaO2/FiO2 ratio when turned to the supine position. This suggests that a short prone period may not be of great benefit to the patient and a longer period of prone positioning may be necessary to allow for postural drainage and the achievement of lasting improvements in the PaO2/FiO2 ratio.

The benefit of longer periods of prone positioning for ARDS patients was noted by Romero et al. (2009) in a pilot study of 15 patients with severe ARDS. The inclusion criteria for the study required a PaO2/FiO2 ratio <100 mmHg at baseline, which is much less than most other reported studies, and indicates the severity of the patients' condition. The study showed a significant increase in PaO2/FiO2 ratio (p < 0·0001) when nursed in a prone position for a total of 55 (±7) h. The prospective study of Johannigman et al. (2000) on 20 ARDS patients also suggested that multiple prone episodes per patient appear to demonstrate a continued improvement in PaO2/FiO2 ratio.

A non-randomized trial by Rossetti et al. (2006) demonstrated a significant correlation between patients' body weight and response to the prone position, with a trend towards those who were heavier responding better. This finding would support the understanding that in the supine position the organs of the abdominal cavity restrict the movement of the diaphragm, thereby reducing residual capacity and closure of the alveoli in dependent zones (Marik, 2001).

It has also been suggested that prolonged episodes of prone positioning improve oxygenation via postural drainage of secretions and reduction in extravascular lung water, which enables re-expansion of the lungs and an improved ventilation/perfusion match. The positive effect of prone positioning on extravascular lung water index (EVLWI) was demonstrated by McAuley et al. (2002), where an initial transient increase was followed by a statistically significant decrease on EVLWI at 18 h.

Incidence of VAP

VAP can be characterized as a pneumonia appearing >48 h after the initiation of mechanical ventilation. The diagnosis of VAP is normally based on radiographic changes showing new infiltrates combined with hyperpyrexia, purulent tracheal aspirations; increased white cell count and microbiology cultures obtained via bronchial lavage (Guerin et al., 2004). The incidence of VAP was reported in three of the included studies. A reduced incidence of VAP in the prone position was noted in two of them and was found to be statistically significant (Guerin et al., 2004; Voggenreiter et al., 2005). The study reported by Fernandez et al. (2008) found that the incidence of VAP actually increased for patients nursed in the prone position, but it should be noted that Fernandez et al. (2008) did not report the diagnostic signs used to identify VAP.

Weaning

The detrimental effects of high FiO2 have been noted to cause oxygen toxicity manifested by damage to the lung parenchyma in the acute phase of lung injury (DuPre and Davis, 1997). The popularity of the prone position increased with the awareness of oxygen toxicity as the immediate rise in PaO2 following pronation allowed for the FiO2 to be reduced. This highlights the importance of implementing weaning within the study design and routine practice.

While the weaning process may initially reduce measures such as PaO2/FiO2 ratio, it may also reduce the risk of complications such as barotraumas, ventilator days, intensive care stay, VAP, sepsis and mortality.

Two studies included in this review reported using tidal volumes in excess of 10 mL/kg (Venet et al., 2001; Mancebo et al., 2006) and two applied ventilation weaning strategies consistent with the ARDS network (2000) recommendations (Voggenreiter et al., 2005; Fernandez et al., 2008). Although the studies reported by Voggenreiter et al. (2005) and Fernandez et al. (2008) noted a reduced mortality rate, the study reported by Mancebo et al. (2006) also suggested a 15% absolute reduction in mortality between the prone- and supine-positioned groups. However, this difference was not statistically significant. These findings would suggest the need for a large-scale study of prone positioning that incorporates a weaning protocol if a significant reduction in mortality is to be demonstrated.

Adverse events

The recording of adverse events in the included studies can be described under four main categories: extubations, dislocation of vascular lines, enteral feeding issues and pressure sores.

Although ventilated patients who are nursed in a supine position should be turned at least every 2 h to prevent pressure sore development and encourage postural drainage (Bryan-Brown and Dracup, 1998), the complex manoeuvres involved in the prone positioning of patients may suggest a higher risk of accidental extubation. In the general ICU population, this is relatively at low risk, approximately 3–4·3/1000 patient days (De Lassence et al., 2002; Garrouste-Orgeas et al., 2010). The studies included in this review also suggest that accidental extubations are a rare adverse event, ranging from 0% to 2·4% (Johannigman et al., 2000; Voggenreiter et al., 2005).

Reported adverse events related to complications of displacement of vascular lines included the interruption of haemofiltration (Mancebo et al., 2006), total removal of central lines and hypotension related to interruption of vasopressor infusions (L’Her et al., 2002), and in the single most severe case the accidental removal of a Swan Ganz catheter (Mancebo et al., 2006).

Enteral feeding issues are only documented in four studies (Johannigman et al., 2000; Gattinoni et al., 2001; L’Her et al., 2002; Mancebo et al., 2006, two of which Johannigman et al., 2000; Mancebo et al., 2006) noted the accidental removal of nasogastric tubes in the prone position. L’Her et al. (2002) reported enteral nutrition intolerance in 25% (n = 13) of patients in prone position with half of these resulting from vomiting, which may increase the risk of aspiration pneumonia. However, no attempt was made to free the abdomen or elevate the head in reverse Trendelenburg position, which may have served to increase abdominal pressure and displace stomach contents. In contrast to this, Voggenreiter et al. (2005) suggested that enteral feeding issues and the incidence of malabsorption occurs equally in patients when nursed in the supine and prone positions.

Pressure sores were noted in nine of the included studies (Johannigman et al., 2000; Gattinoni et al., 2001; L’Her et al., 2002; McAuley et al., 2002; Guerin et al., 2004; Voggenreiter et al., 2005; Mancebo et al., 2006; Fernandez et al., 2008; Romero et al., 2009). However, it should be noted that the complex pathology of critically ill-ventilated patients often means that they are more susceptible to the development of pressure sores (Shahin et al., 2008). Voggenreiter et al. (2005) noted a 63% and 90% incidence of pressure sores in the supine and prone groups, respectively. However, the skin lesions were not documented at the outset of the study, which could suggest that they may have been related to the patients' underlying condition or injury rather than the position they were nursed in.

Mortality

Documented mortality rates of patients nursed in the prone position vary considerably within the included studies, ranging from 5% (Voggenreiter et al., 2005) to 49% (L’Her et al., 2002). As a result of the study design and difficulties of recruitment, many of the included studies did not reveal any reduction in mortality, which could be conclusively attributed to the prone positioning intervention. Gattinoni et al. (2003) performed a retrospective analysis of high-risk patients and found no difference in overall mortality for patients nursed in the prone position, but did find a survival advantage at day 10, suggesting that multiple prone episodes may be beneficial for high risk and patients with severe ARDS. Mancebo et al. (2006) reported a 15% absolute decrease in ICU mortality for patients nursed in the prone position, and Fernandez et al. (2008) found a 15% absolute increase in 60-day survival rates. However, it should be noted that these findings were not statistically significant. The retrospective review reported by Davis et al. (2007) did find a statistically significant difference in mortality because of positioning. However, a large number of patients in the supine position group had severe closed head injury and when these were excluded from the analysis, the difference between the prone and supine groups were not significant.

DISCUSSION AND IMPLICATIONS FOR NURSING PRACTICE

It is acknowledged that a narrative review is limited by the often subjective nature of interpretation, nonetheless some important issues for ICU nurses caring for patients with ALI or ARDS have been highlighted, which, in the absence of conclusive guidelines for practice, may help them in contributing to the management plan for their patients or in research design. The search strategy limited studies to those published since 2000; however, the development of health technologies in the intensive care environment means that clinical research findings can be very quickly superseded by technological innovations. It may also have been helpful to contact the reviewers (Bloomfield et al., 2009) who have registered a review protocol on the Cochrane database, but this review was particularly interested in interpreting the nursing implications of prone positioning of ventilated patients.

The studies included in this narrative review did not show conclusive evidence of the clinical benefits of nursing patients with ALI or ARDS in the prone position. Some of this inconsistency may be because of differences in study design, difficulties with recruitment or an absence of practical detail in the reporting of the studies. For example, within the 14 included studies, the operational definition of prone positioning was largely unclear. It is a general recommendation that ventilated patients should be nursed with the head of bed elevated to 30–45° to reduce the risk of gastroesophageal reflux and nosocomial pneumonia (Ibanez et al., 1992). However, only Guerin et al. (2004) reported detail of a prone position incorporating a reverse Trendelenburg position, although the degree of incline was not documented. It was also reported in this study that the degree of elevation for the patients allocated to the control group was not accurately measured during the trial.

Nonetheless, it does appear from this review that prone positioning has an impact on the PaO2/FiO2 ratio. The research also seems to suggest that a short prone period may not be of great benefit to the patient and a longer period of prone positioning may be necessary to allow for postural drainage and achieve improvements in the PaO2/FiO2 ratio. Postural drainage may also correlate with a reduction in the incidence of VAP, which itself may have a profound impact on many aspects of critical care outcomes and an overall reduction in ICU mortality rates (Safdar et al., 2005). In the day-to-day care and management of ventilated patients, an important aspect of nursing practice relates to the weaning of the patient from the ventilator. Not all the studies applied a weaning protocol to the patients, only weaning the fraction of inspired oxygen (FiO2). Although this would demonstrate an immediate effect of the positional intervention, it may not improve the overall patient outcome as weaning should be a continuous process. The evidence reviewed here also suggests that the effect of prone positioning on mortality in patients with ALI or ARDS is not conclusive. However, considering the critical status of these patients, any improvement in their condition could perhaps be regarded as an achievement.

Despite these suggested clinical benefits, there is little evidence to suggest whether all ventilated patients may derive benefit from prone positioning. There is also a lack of conclusive evidence to indicate whether prone positioning should be routinely practiced from the outset of a patient's stay in ICU or whether this should only be an intervention given to high-risk patients, or indeed should only be implemented once a patient's condition deteriorates. If any clinical benefits can be conclusively associated with prone positioning, then it is arguable that this intervention should be offered to all ventilated ICU patients. However, this would inevitably have considerable cost and resource implications, so it may be useful to all those responsible for providing care in the ICU if future research in this area were able to explore both the clinical and economic impacts of prone positioning.

In relation to improving overall respiratory function of this critically ill patient group, it seems that one of the physical consequences of prone positioning is the reduction of abdominal pressure on the thorax. The ‘free-abdomen’ approach can be achieved by the use of specific prone position devices or the adjunct of pillow placement under the shoulders and pelvis (Rowe, 2004). However, only Johannigman et al. (2000) reported the use of a specialist proning device (Vollman Prone Positioner; Hill-ROM, Batesville, IN, USA), and none of the studies documented the use of cushions. Hering et al. (2001) and Rossetti et al. (2006) did report that no cushions or measures were taken to alleviate positional restrictions of the abdomen. This means that ICU nurses do not have clear guidance as to how best to achieve a ‘free abdomen’ in patients who they position prone.

In relation to the practicalities of moving ventilated patients, all ICU nurses are aware that extra care must be taken when positioning intubated patients and although the evidence about adverse events associated with prone positioning is again inconclusive, it may be advisable for nurses to take additional precautions when moving patients from a supine to a prone position. The simple addition of extensions to all vascular lines may help ICU nurses to minimize the risk of adverse events occurring in these patients who are already severely compromised by their condition.

The potential benefits of nursing a critically ill patient in the prone position should always outweigh the potential risk in moving the patient, so ICU nurses need to ensure that enough staff with appropriate experience are involved in any manoeuvring of a ventilated patient into a prone position. In terms of the everyday practicalities of nursing patients in the prone position, it would be useful to have guidance on how nurses can manoeuvre their patients in a way that carries maximum clinical benefit and minimum patient risk.

CONCLUSIONS

This review suggests that prone positioning has some benefits for patients suffering from ALI and ARDS and is best applied in multiple episodes for long periods, using a reverse Trendelenburg position with a free abdomen. However, the variation in design of the reviewed studies means that the evidence is not conclusive, and it is not possible to draw robust conclusions about the effectiveness of prone positioning on clinical outcomes in this patient group. The available evidence does not provide guidelines as to when instigation of the prone position may be most beneficial to patients. It should also be noted that the reviewed evidence applied manual proning procedures and that there is, as yet, no comparative research about automated proning devices.

It would appear that more research is needed in this field, and future studies may benefit from a multicenter and multi-professional approach, together with clear operational definitions of both prone and supine conditions. If there are any benefits to be gained from nursing severely ill patients in the prone position, it is important that ICU nurses understand both the clinical and practical dimensions of this intervention, to ensure they are able to deliver the best possible care.

ACKNOWLEDGEMENTS

This review was carried out as part of the Master of Research in Health Sciences where AW is a full-time postgraduate student, funded by the NIHR Clinical Academic Training programme.

WHAT IS KNOWN ABOUT THIS TOPIC

  • Mortality rates for ventilated patients with ALI or ARDS are very high, which presents a challenge to ICU nurses caring for them.
  • A growing body of evidence suggests that nursing these patients in the prone position may have a beneficial impact on clinical outcomes.
  • Awareness of this evidence can empower ICU nurses to make informed clinical decisions about positioning their patients.

WHAT THIS PAPER ADDS

  • This review of available evidence suggests that the clinical benefits of prone positioning are not yet conclusive.
  • Systematic review of the evidence or large-scale clinical trials of prone positioning in ICU are needed to inform guidelines for practice.
  • In the absence of robust guidelines, ICU nurses need to integrate their clinical expertise with the best available evidence in order to position their patients appropriately and to provide comfort for the patient and their family.

Ancillary