Authors: Andrea Bruni, Eugenio Garofalo, Laura Grande, Gaetano Auletta, Davide Cubello, Manfredi Greco, Nicola Lombardo, Pietro Garieri, Anna Papaleo, Patrizia Doldo, Rocco Spagnuolo, Federico Longhini
Categories: Review Article, Enteral nutrition, Nurse, Prone position, Regurgitation, Residual gastric volume, Vomiting
Source: Intensive & Critical Care Nursing
Early enteral nutrition (EN) and prone position may both improve the outcome of patients affected by moderate to severe Acute Respiratory Distress Syndrome. Recent guidelines suggest to administer early EN also during prone position. However, EN intolerance, such as high residual gastric volumes, regurgitation or vomiting, may occur during pronation.
This systematic review aims to assess the occurrence of high residual gastric volume, regurgitation or vomiting episodes, that can be encountered in patients receiving EN during prone position.
We have conducted a systematic review. We queried three scientific databases (MEDLINE, EMBASE and CINAHL) from inception until November 19, 2019 without language restrictions, using keywords and related MeSH terms. All relevant articles enrolling adult patients receiving invasive mechanical ventilation and evaluating the use of early EN during prone position were included.
From 111 records obtained, we included six studies. All studies but one reported no differences with respect to gastric residual volumes between supine and prone positions. A 24-hours EN administration protocol seems to be better, as compared to an 18-hours feeding protocol. The need to stop EN and vomiting episodes were higher during prone position, although the rate of high gastric volume was similar between supine and prone positions. Ventilator associated pneumonia, lengths of stay and mortalities were similar between supine and prone positions. Only one study reported lower mortality in patients receiving EN throughout the entire day, as compared to an 18-hours administration protocol.
Protocols should be followed by healthcare providers in order to increase the enteral feeding volume, while avoiding EN intolerance (such as EN stops, high residual volume, regurgitation and vomiting).
Keywords: Enteral nutrition, Nurse, Prone position, Regurgitation, Residual gastric volume, Vomiting
Several decades ago, prone positioning was proposed in patients with hypoxaemic acute respiratory failure to improve gas exchange (Piehl and Brown, 1976). When applied early and for prolonged (>16 hours) sessions, prone position reduces the 28-days mortality in patients affected by Acute Respiratory Distress Syndrome (ARDS) (Guerin et al., 2013). Although a life saving procedure, prone positioning may be affected by some complications, such as nasogastric feeding tube, central venous catheters or orotracheal tube dislodgements or increased risk for pressure lesions (Lucchini et al., 2018).
According to the recent guidelines released by the European Society for Clinical Nutrition and Metabolism (ESPEN), enteral nutrition (EN) should be considered, in the absence of contraindications, as early as possible (within the first 48 hours from Intensive Care Unit (ICU) admission) in all critically ill patients, even in those managed in the prone position (Singer et al., 2019). In addition, also the European Society of Intensive Care Medicine (ESICM) recommends that early EN should not be delayed solely because of prone positioning. Furthermore, ESICM guidelines suggest considering the early use of prokinetics, followed by post-pyloric feeding, in case of persisting gastric retention (Reintam Blaser et al., 2017). Indeed, early EN reduces the ICU and hospital mortality (Artinian et al., 2006), length of stay (Marik and Zaloga, 2001) and infectious complications (Marik and Zaloga, 2001).
Deep sedation, septic shock, haemodynamic impairment, supine positioning without head elevation and elevated intra-abdominal pressures are common in mechanically ventilated ARDS patients during prone position; of note, these factors may alter gastric motility and delay emptying, leading to high residual gastric volume, regurgitation or vomiting episodes (Linn et al., 2015). Thus, nursing care is fundamental to monitor and recognise EN intolerance, in order to implement interventions and to improve enteral feeding. We have therefore designed this systematic review of the literature to assess issues of interest that nurses could face in patients receiving early EN during prone positioning.
We have conducted this systematic review in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement (Liberati et al., 2009).
Three scientific databases (MEDLINE, EMBASE and CINAHL) were searched from inception until November 19, 2019 without language restrictions, using keywords and related MeSH terms. The search strategy is detailed in the Electronic Supplemental Material (ESM). Controlled vocabulary terms, text words and keywords were variably combined. Blocks of terms per concept were created.
We included all relevant articles enrolling adult patients receiving invasive mechanical ventilation and evaluating the use of early EN during prone position. We excluded papers with one or more of the following published in languages other than English, Italian, French or Spanish; case report or series; review; systematic reviews or meta-analysis. References of included papers and review articles were also examined to identify additional studies missed during the primary search.
We recorded all the following outcomes, considered issues of potential interest to residual gastric volumes, EN fed volume, need to stop the EN, rate of episodes of vomiting or regurgitation, rate of ventilator-associated pneumonia (VAP) or secondary infections, ICU and hospital length of stay and mortality. For all measured outcomes, we considered the definition provided in every single study included in the review.
Two authors (AB and EG) independently screened the retrieved records for exclusion by title and, then, by abstract, according to the inclusion/exclusion criteria. Full-texts of possible relevant articles were assessed for inclusion/exclusion. Data were independently extracted by two authors (AB and EG), and collected through a dedicated electronic form, specifically designed for the present review, on Microsoft Excel (Microsoft Corporation, Redmond, WA, USA). All disagreements were discussed and, if necessary, resolved by a third author (FL) at every step of the process.
Data were extracted as reported in the original manuscripts. The methodological quality of included studies was assessed using Review Manager software (RevMan 5.3; Nordic Cochrane Centre, Cochrane Collaboration, Copenhagen, Denmark). We evaluated all studies for randomised sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other bias. Randomised and nonrandomised crossover studies were assessed according to a modified checklist (Ding et al., 2015).
Descriptive statistics of individual studies used different statistical indicators for central tendency and variability, such as means and standard deviations (SD), whereas absolute and relative frequencies were adopted for qualitative variables. To show one single indicator for the quantitative variables we collected, means (SD) or medians and interquartile ranges [IQR] were used, as appropriate (Messina et al., 2017, Messina et al., 2018). The scarcity of data published in the literature and low study quality, prevented the possibility of conducting a pooled data analysis for outcomes.
The study selection flowchart is reported in Fig. 1 . After launching the search strategy, 111 titles were identified. Fifteen duplicated records and 89 citations were excluded on the initial title and abstract screen. After examination of the remaining seven full text, six studies have been included in the systematic review (Lucchini et al., 2017, Reignier et al., 2010, Reignier et al., 2004, Saez dela Fuente et al., 2016, Sams et al., 2012, van der Voort and Zandstra, 2001). One article was excluded because it was a narrative review. Risk of bias assessment is depicted in Fig. 2 .
Fig. 1 Flow Diagram. Flow diagram of studies screening and selection according to the Preferred Reporting Items for Systematic review and Meta-Analysis (PRISMA) recommendations.
Fig. 2 Risk of bias assessment. The risk of bias has been assessed for all included studies. All studies suffer of a high (red) or intermediate (white) risk for considered biases. One only study shows a low (green) risk of selection bias. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
All included studies are single centred (Lucchini et al., 2017, Reignier et al., 2010, Reignier et al., 2004, Saez dela Fuente et al., 2016, Sams et al., 2012, van der Voort and Zandstra, 2001). Design and characteristics of included studies are reported in Table 1 .
Overall, the six studies included 241 patients, with a median of 30 [21–62] patients per study. The median age was 55 [50–61] years. Of the 241 patients, 96 were admitted for pneumonia, 65 for ARDS, 23 for sepsis/septic shock, 21 for trauma, 13 for cardiogenic shock and 23 for other miscellaneous reasons. All studies but one (Lucchini et al., 2017) reported the gender; 149 patients were males and 67 females. Three studies did not report any severity score at ICU admission (Lucchini et al., 2017, Saez dela Fuente et al., 2016, Sams et al., 2012). One study reported the average APACHE score, i.e. 25.5 (8.98) (van der Voort and Zandstra, 2001), one study the Simplified Acute Physiology Score II (SAPS-II), i.e. 52 (14) and 52 (30) in the two groups of randomised patients (Reignier et al., 2004) and one study both SAPS-II (i.e. 52 (15) and 55 (16) in the two studied population) and the Sequential Organ Failure Assessment (SOFA) score, (10 (3) and 9 (4), respectively) (Reignier et al., 2010).
Van der Voort and Zandstra prospectively assessed the tolerance of EN during prone position, as compared to supine position. EN was initiated within 24 hours of admission, at an increasing rate up to 80ml/hour. All patients were studied for 6 hours in the prone position and 6 hours in the supine position, with a 30° head elevation in both positions (van der Voort and Zandstra, 2001).
Reignier and colleagues prospectively assessed the tolerance of EN during prone position in 71 patients with severe hypoxemic Acute Respiratory Failure. EN was started at 30ml/hour and daily increased by 30ml/hour, in order to administer 500ml of feeding on the first day, 1000ml on the second, 1500ml on the third and 2000ml on the fourth and fifth days. EN was continuously delivered from 6pm to noon, while no discontinuation was applied in patients treated with insulin (Reignier et al., 2004).
Another studied by Reignier (Reignier et al., 2010) compared two different strategies to provide early EN in patients in prone position. This before-after study compared a strategy (control group) previously described (see above) (Reignier et al., 2004), with a new protocol consisting of EN delivery over 24 hours, starting at 25ml/hour and increased by 25ml/hour every 6 hours up to 80ml/hour, in absence of intolerance (Reignier et al., 2010).
In ARDS patients during prone positioning, Sams and colleagues assessed if the incidence of micro-aspiration, detected by pepsin A assay in the tracheal aspirate was reduced by the insertion of post-pyloric tube, as opposed to nasogastric feeding tube (Sams et al., 2012).
In 34 patients, Saez de la Fuente and colleagues assessed tolerance and safety (i.e. high gastric residual events, vomiting episodes or regurgitation) of EN in prone position, as compared to the supine position. EN was delivered over 24 hours, adjusting the rate to gradually achieve an energy target of 25kcal/kg/day, in steps of 25% over the first four days (Saez dela Fuente et al., 2016).
Lastly, Lucchini and colleagues retrospectively compared the gastric residual volume in 25 ARDS patients receiving EN during supine and prone position. EN delivery rate was adjusted according to a protocol based on the gastric residual volume checks (every 6 hours), and modifying the rate of infusion at steps of 21ml/hour (Lucchini et al., 2017).
Residual gastric volumes are reported by five studies (Lucchini et al., 2017, Reignier et al., 2010, Reignier et al., 2004, Saez dela Fuente et al., 2016, van der Voort and Zandstra, 2001). Single studies results are shown in Table 2 . Van der Voort and colleagues reported no differences of gastric residual volume between prone and supine position after 3 and 6 hours of feeding (van der Voort and Zandstra, 2001). On the contrary, Reignier (Reignier et al., 2004) described a small, though significant difference in the residual gastric volume between prone and supine positions at day 1, 2 and 4; of note, the amount of enteral feeding administered during prone position was lower, as compared to supine position (Reignier et al., 2004). The rate of events of excessive residual gastric volume was similar between patients’ groups in four studies (Lucchini et al., 2017, Reignier et al., 2010, Saez dela Fuente et al., 2016, van der Voort and Zandstra, 2001).
In the study by Reignier (Reignier et al., 2010), the prolongation of EN delivery over 24 hours significantly increased the daily administered feed volume, without any impact on the gastric residue. Finally, both Saez de la Fuente and colleagues and Lucchini and colleagues did not report differences in the residual gastric volumes between prone and supine positions (Lucchini et al., 2017, Saez dela Fuente et al., 2016); furthermore, in the latter study, the EN fed volume was similar between the two conditions (Lucchini et al., 2017).
Early EN intolerance events during prone position were reported as the need of EN discontinuation, high residual gastric volumes, vomiting or regurgitation episodes.
The need to stop EN was reported by two studies (Lucchini et al., 2017, Reignier et al., 2004). Reignier and colleagues reported a higher rate of EN discontinuation during prone position (82% of patients, with a total of 63 times of discontinuation), as opposed to supine position (49% of patients, with a total of 42 times of discontinuation) (Reignier et al., 2004). In the study by Lucchini and colleagues, EN was stopped in 9.2% of assessments during supine position, while in 6.8% during prone position; noteworthy, only once the suspension was attributable to an excessive residual gastric volume (>500 ml) (Lucchini et al., 2017).
Four studies have reported the number of patients showing high residual gastric volumes (Lucchini et al., 2017, Reignier et al., 2010, Saez dela Fuente et al., 2016, van der Voort and Zandstra, 2001). Noteworthy, the definitions of high residual volumes were different among studies. Van der Voort and Zandstra defined the residual gastric volumes as “high” when >150ml after 6 hours of nutrition (van der Voort and Zandstra, 2001), Reignier if >250ml (Reignier et al., 2010), Lucchini if >300ml (Lucchini et al., 2017) and Saez de la Fuente if >500ml (Saez dela Fuente et al., 2016). In the study by van der Voort and Zandstra, six patients showed a residual gastric volume >150ml in both supine and prone position, while one patient showed a high volume >150ml only during prone position (van der Voort and Zandstra, 2001). In the study by Reignier and colleagues (2010), intolerance to EN occurred in 71% of control patients and 63% of intervention patients (p = 0.5). Furthermore, the presence of high residual volume was similar between intervention (58%) and control (59%) groups (p = 0.6) (Reignier et al., 2010). Lucchini reported that only one patient (4%) suffered of high gastric residual volume during both supine and prone positions (Lucchini et al., 2017). Finally, Saez de la Fuente and colleagues did not report differences in the number of high gastric residual events per day between supine and prone positions (0.06 vs. 0.09; p=0.39) (Saez dela Fuente et al., 2016).
Vomiting or regurgitation events were reported by four studies (Reignier et al., 2010, Reignier et al., 2004, Saez dela Fuente et al., 2016, van der Voort and Zandstra, 2001). Van der Voort and Zandstra reported only one (5%) patient vomiting during prone positioning, while none during supine. Reignier and colleagues (2004) reported an increased risk for vomiting during prone position (p<0.001; relative risk, 2.5; 95% confidence interval, 1.5–4.0). In particular, 30 episodes of vomiting were reported during the 218 periods of the prone position, while 26 episodes during the 462 supine position periods (Reignier et al., 2004). In another study by Reignier (Reignier et al., 2010), vomiting occurred in all patients receiving the 18-hour feeding protocol, while in 9 out of 12 patients receiving 24-hour feeding protocol (Reignier et al., 2010). In addition, Saez de la Fuente and colleagues reported no differences between supine and prone positions with regard to vomiting (0.016 vs 0.03; p=0.53) and diet regurgitation events per day (0 vs 0.04; p=0.051) (Saez dela Fuente et al., 2016). Finally, Sams and colleagues showed that the insertion of post-pyloric feeding tube may provide a slight, though not significant, protective effect for aspiration risk in prone position, when compared to nasogastric feed (odds ratio 0.778; 95% confidence 0.09–6.98) (Sams et al., 2012).
Only the two studies by Reignier and colleagues assessed some clinical outcomes between treatments (Reignier et al., 2010, Reignier et al., 2004). In the first trial, the rate of ventilator-associated pneumonia was similar between supine (24%) and prone position (35%) (Reignier et al., 2004). Mortality was also similar between groups (24% vs. 35%, respectively) (Reignier et al., 2004).
In the second trial (Reignier et al., 2010), 10 (29%) controls and 9 (24%) interventional patients developed a ventilator-associated pneumonia (p=0.58). The incidence of ventilator-associated pneumonia was 2.4 episodes every 100 patient-days of intubation in controls, while 1.6 episodes in the intervention group. Furthermore, secondary infections (such as urinary tract infections, bacteremia, catheter-related blood stream infections) occurred in 16 (47%) patients in the control group and 16 (42%) in the intervention group (p = 0.81). Reignier and colleagues did not find difference with respect to ICU length of stay (20 (13) in the control vs. 23 (13) days in the interventional group; p=0.56), and hospital length of stay (28 (21) in control vs. 34 (29) days in the interventional group; p=0.31). Finally, patients in the interventional group had significantly lower rates of ICU (28%) and hospital (29%) mortality, compared to the control group (53%, p=0.04 and 29%, p=0.009, respectively) (Reignier et al., 2010).
The current literature on administration of EN during prone position is scanty. Despite the low number of studies, administration of EN during prone positioning seems to not increase the gastric residual volumes to a clinically relevant extent (Lucchini et al., 2017, Reignier et al., 2010, Reignier et al., 2004, Saez dela Fuente et al., 2016, van der Voort and Zandstra, 2001), although only one study reported an increased need to stop EN and rate of vomiting in prone position (Reignier et al., 2004). On the contrary, other studies did not report any increased risk for regurgitation, vomiting or pneumonia during prone position (Reignier et al., 2010, Saez dela Fuente et al., 2016, van der Voort and Zandstra, 2001). Furthermore, protocols including strategies to increase the EN tolerance (head-of-bed elevation, use of prokinetic agents, continuous administration over 24 hours) may be effective to augment the feeding volume in patients during prone position (Reignier et al., 2010). Finally, the rate of ventilator associated pneumonia was similar between patients receiving EN during supine or prone position (Reignier et al., 2004).
The recent ESICM guidelines suggest that early EN should be preferred to delayed EN, because it reduces the risk of infection (Reintam Blaser et al., 2017). Furthermore, early EN: 1) should be preferred to early parenteral nutrition, 2) should be started at low doses as soon as the shock is controlled with fluids and vasopressors/inotropes, 3) should be administered in patients with controlled hypoxaemia and compensated or permissive hypercapnia and acidosis and 4) should be assured also in case of deep sedation and infusion of neuromuscular blocking agents (Reintam Blaser et al., 2017). In keeping with the ESICM guidelines (Reintam Blaser et al., 2017), the recent ESPEN guidelines strongly suggest that early EN should also be performed in patients managed in the prone position (Singer et al., 2019).
Based on the evidence from the literature and, finally, on the indications from guidelines (Reintam Blaser et al., 2017) (Singer et al., 2019), there is a growing role for early EN even during prone positioning in ARDS patients. Of note, when applied in the early phase of moderate to severe ARDS, prolonged (>16 hours) prone positioning sessions halved the 28-day mortality (Guerin et al., 2013). However, gastric emptying is delayed in up to 60% of critically ill patients; this is particularly true in patients during prone position, with increased risk for EN intolerance, regurgitation or vomiting (Linn et al., 2015). In this systematic review, only one study has reported an increased risk for EN intolerance and vomiting, deeming necessary to stop EN (Reignier et al., 2004). In the case medical or nursing staff face with EN intolerance, some corrective and preventive actions can be implemented. For instance, the use of different and predefined nutritional protocols, including the use of prokinetics agents, post-pyloric feeding, head-of-bed elevation and continuous administration over 24 hours, may increase EN tolerance during prone position (Reignier et al., 2010, Sams et al., 2012). Furthermore, a closed monitoring of gastric residual volumes and adjustment of the rate of enteral feeding is fundamental to increase EN tolerance (Lucchini et al., 2017).
In this scenario, intensive care nurses should be aware of the higher risk for regurgitation and vomiting during prone position (Reignier et al., 2004); therefore, they should be properly trained to keep the patient with head-of-bed elevation (Reignier et al., 2010), to monitor enteral feeding and to detect the potential signs of EN intolerance (Lucchini et al., 2017, Reignier et al., 2010). The presence of a predefined protocol for EN management should be implemented to guide nurses in their daily activity, to administer higher enteral feeding volumes and to avoid EN stop, regurgitation or vomiting episodes (Reignier et al., 2010).
Although this systematic review is contemporary, it suffers from some important limitations. First of all, there is a clear lack of prospective randomised controlled trials, properly designed to assess differences in major clinical outcome (such as infections, lengths of stay, mortality). Furthermore, the heterogeneity of study designs, outcomes definition and assessment, together with a low yield of the literature search, preclude quantitative assessment of pooled study results. In addition, the quality of the included studies is quite low, mainly due to the retrospective or before-after design. Finally, we could not provide data regarding the real-life use of EN during prone position, due to the lack of targeted observational studies. Therefore, it is desirable that future prospective and randomised studies will 1) the real use of EN during prone position, 2) the development of a properly designed protocol to deliver, monitor and guide nurses in EN administration during prone position; 3) the incidence of complications or EN intolerance events during prone position; 4) if the early administration of EN during prone position may improve major clinical outcomes of patients. All of the above is important to confirm the safety and tolerability of EN during prone position, and to provide solid evidence and strong recommendations in future guidelines.
Although the literature is poor and scanty, EN during prone positioning is currently recommended by the guidelines. Both the medical and nurse staff should follow protocols in order to increase the enteral feeding volume, while avoiding EN intolerance (such as EN stops, high residual volume, regurgitation and vomiting). The literature currently lacks of definitive data from randomised controlled trials designed to assess if the administration of early EN during prone position may influence major clinical outcomes.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
The following are the Supplementary data to this