Early, adequate enteral nutrition (EN) is associated with improved patient outcome and reduced duration of mechanical ventilation and length of hospital stay (Heyland et al, 2011; Compher et al, 2017). Meeting the nutritional needs of intensive care patients is complex and multifactorial: patients have increased protein demand, yet there may be inadequate protein provision in standard feeds, and patients may experience delays in feed initiation secondary to interventions and access (Taylor et al, 2016). International research of 158 intensive care units (ICU) found that about 50% of energy prescriptions and 60% of protein prescriptions are being met (Cahill et al, 2010; Cahill et al, 2014). It is, therefore, vital to ensure safe tube placement and securement, because these are prerequisites for effective EN delivery in critically unwell patients (Taylor et al, 2015).
Up to 58% of patients admitted to ICU require nasogastric (NG) or nasointestinal (NI) feeding (Sharifi et al, 2011). When secured with tape, feeding tubes are easily dislodged. Inadvertent tube loss, slippage and patient removal occurred in 54% of ICU-placed tubes, of which 44% required replacement (Taylor et al, 2014). Preventing tube loss can reduce the associated risks with tube reinsertion, such as repeated lung placement, excessive X-ray exposure and delays to feeding or administration of medication (Bechtold et al, 2014; Brazier et al, 2017). Securement techniques include adhesive tape, hand mittens or additional nurses to supervise patients who are likely to remove feeding tubes (Brazier et al, 2017). An alternative is to use a nasal bridle as a retention device. This is looped behind the vomer bone and externally clipped to the feeding tube, and the tube is then secured using tape or monofilament. Existing research, including a meta-analysis, found that replacing tape securement with a nasal bridle reduced inadvertent tube loss from 40–53% to 9–20% of patients (Bechtold et al, 2014; Taylor et al, 2018).
Literature review
A review of critically ill patients fed via an NI tube identified a higher incidence of tube removal of those secured with tape compared with those secured with nasal bridles, 48% versus 13% respectively, P=0.0006 (Seder et al, 2010). The systematic review and meta-analysis included six papers, focusing on tube dislodgement. Of the six papers, five were prospective cohort studies and one a randomised controlled trial (RCT). Of the patients included in the meta-analysis (n=544) the results favoured the use of a bridle to secure feeding tubes, which were more effective than using adhesive tape alone (Bechtold et al, 2014). Furthermore, the RCT found a higher percentage of the calorie target was delivered to patients whose tubes were secured with bridles compared with tape securement, 79% versus 62% respectively, P=0.016 (Seder et al, 2010).
In a small, retrospective service evaluation of critically unwell NI-fed patients (n=50), nasal bridles significantly increased protein delivery in those who had a bridle compared with those who did not (75% versus 55%, P=0.033) (Allan et al, 2018). There is growing evidence across the ICU, stroke and burns population linking the use of nasal bridles with reduced hours of missed feeding and improved nutritional (calorie and protein) delivery.
A recent study in stroke patients found that 73% of tubes were inadvertently removed (patient: 64%; slippage: 9%). Delays to initiating feed were long (8.5 hours) and related to 65% of tubes requiring X-ray confirmation of positioning (Brazier et al, 2017).
In burns patients, the use of nasal bridles resulted in 80% fewer dislodgement events (P=0.005) and fewer hours of missed feeding time (P=0.05) compared with tubes secured with tape (Li et al, 2018). Placing a nasal bridle was also associated with fewer tubes being inserted and reduced delays to feeding (Li et al, 2018).
Local practice in Southmead Hospital ICU between 2015 and 2017 specified that a bridle should be used when the following was the case: tape securement failed to retain the initial feeding tube, the tube was a vital drug route, or tube placement carried high risk.
The aim of the review reported in this article was to audit nasal bridle efficacy and current nasal bridle practice (tape) in one ICU. The primary outcome was to measure tube loss. The secondary outcome was to measure the duration of tube use, comparing tubes secured with a nasal bridle with those secured by tape on day 1 in enterally fed patients in ICU.
Methodology
Study design
We conducted an observational prospective audit of current practice, collecting data on all patients newly admitted to Southmead Hospital ICU between October and December 2017 preparatory to reviewing the bridle policy in January 2018. Data collection included nasal bridle insertion date, the number of tubes inserted and duration of tube use, and was undertaken mainly by the lead author (KA), assisted by co-authors RC and ST.
Audit participants
All ICU patients with NG or NI tubes in place were screened daily by the authors during the audit period to measure the number of tubes inserted and the method of tube securement. The tubes were initially secured with tape (as standard), and a nasal bridle was subsequently used in cases where the tube had been removed once or placed alongside the first tube, if clinically indicated. Nasal bridles were routinely placed by nurses and recorded on the ICU observation charts. All NI tubes were sited by the dietitians and had routine bridle securement. All nurses and dietitians placing nasal bridles had received appropriate training in accordance with Trust policy.
Patients were excluded from the audit if the nutritional delivery route was oral, orogastric, orointestinal (where the tube is placed via the mouth), or if the administration route was parenteral (nutrition delivered via the vein).
Ethical considerations
This observational audit did not require ethical approval because the interventions were clinically indicated and data collection was part of routine dietetic records. Consent is difficult to obtain from ICU patients due to their reduced consciousness, ventilation, sedation and intubation status, and patients often lack capacity to make decisions while acutely unwell. As a result, feeding tubes and nasal bridles are placed in the patients' best interests in accordance with the Mental Capacity Act 2005 and in line with the ICU nasal bridle standard operating procedure document. Data were stored within a password-protected file, to which only the authors had access, and data treated in line with the Data Protection Act 2018. No financial support was given to undertake this project.
Analysis
Statistical analysis was undertaken using R Studio v1.1.383 software (ST/DT). The analyses included testing the normality of the data, and univariate and regression analysis was carried out to determine the effect of bridle securement on tube loss, duration of use and achievement of tube redundancy, ie when the tube was no longer required.
Findings
A total of 109 patients were recruited, with 205 tubes placed over the 3-month period. The number of patients in each group was not equal: n=63 unbridled versus n=46 bridled.
Statistical analysis
A power calculation was performed, which showed that 102 patients (n=51 unbridled; n=51 bridled) were required for the study to have a power of 90% to detect a difference at a significance level of 0.05, assuming that inadvertent tube loss reduces from 50% to 20% when moving from tape to bridle securement. A power calculation using R studio (power.prop. test) was performed, which determined a significance level of P=0.05 for tube loss. Two-tailed, reverse power calculations using the actual study group size and percentages of tubes lost or reaching redundancy had a power of 99% and 92% respectively. The Shapiro-Wilk test determined that most parameters were not normally distributed, and so they were analysed using the median interquartile range (IQR), Wilcoxon signed-rank tests for continuous data, and the Fisher's exact test for categorical data.
Because the tape group had a higher proportion of NG versus NI tubes than the bridle group, the patients' age, sex, disease state, chronic health evaluation score (APACHE II score) (Knaus et al, 1985), state of consciousness and the route of administration were entered into linear or logistic regression models to test for independent statistical associations. Variables with a P<0.2 were retained and re-tested in regression model 2 to identify the most strongly associated variables. Model 3 retained model 2 variables with a P value less than 0.1 and added the ‘bridle’ variable to determine whether it was independently associated with outcomes.
Patient demographics
Over the 3 months, 77 nasal bridles were placed with the 109 patients who were recruited. On day 1 only 42% of patients (n=46) had a bridle fitted, with most patients being sedated or unconscious during insertion. Routine practice during the study period was to secure all NI tubes with bridles on day 1, unless contraindicated. The bridled group had more NI tubes than NG tubes compared with patients without bridles due to the dietitians placing bridles as standard alongside NI tube placements (Table 1).
| Parameter | No bridle n=63 (57.8%) | Bridle n=46 (42.2%) | P-value | |
|---|---|---|---|---|
| Median (IQR) | Median (IQR) | |||
| Age (years) | 61 (46–73.5) | 62 (47–71.8) | 0.95 | |
| Sex | % male | 30 (76.2) | 28 (69.9) | 0.18 |
| % female | 33 (23.8) | 18 (29.1) | ||
| APACHE 2* | 15 (7.5–19.5) | 13 (0–18.5) | 0.26 | |
| n (%) | n (%) | |||
| Disease (%) | Medical | 24 (38.1) | 17 (37.0) | 0.67 |
| Neurosurgical (non-trauma) | 19 (30.2) | 15 (32.6) | ||
| Surgical (non-trauma) | 10 (15.9) | 4 (8.7) | ||
| Trauma | 10 (15.9) | 10 (21.7} | ||
| Consciousness (%) | Awake | 17 (27) | 10 (21.7) | 0.65 |
| Sedated/unconscious | 46 (73) | 36 (78.3) | ||
| Route | Gastric | 63 (100) | 46 (93.5) | 0.034 |
| Intestinal | 0 (0) | 3 (6.5) | ||
| Tube size (FG) | 8 | 3 (4.8) | 2 (4.3) | 0.60 |
| 10 | 1 (1.6) | 3 (6.5) | ||
| 12 | 57 (90.5) | 39 (84.8) | ||
| 14 | 2 (3.2) | 2 (4.3) | ||
FG = French gauge, IQR = interquartile range
Outcomes
Tube securement with a bridle is associated with a lower likelihood of tubes being removed due slippages or patients pulling at it. The tubes therefore reach natural redundancy, for example, when the patient can eat and drink.
Primary outcomes: compared with tubes secured with a bridle at day 1, unbridled tubes were more likely to be prematurely lost (17.4% versus 57.1% respectively, P<0.0001) (Figure 1).
Secondary outcomes: tube securement at day 1 with a bridle, compared with tape alone, was associated with a longer duration of a tube remaining in situ (median [IQR] 10 [7–17] versus 5 [2–10] days respectively, P<0.0001) and enteral nutrition (12 [8–17] versus 5 [2–10] days respectively, P<0.0001). In addition, there was an increased likelihood that the tube would reach redundancy because it was no longer required (36.2% versus 68.4%, P<0.003). Use of bridles was associated with reduced patient removal (13% versus 41.3%) and slippage (2.2% versus 9.5%), but an increase in the tube being replaced by oral nutrition (39.1% versus 17.5%) and long-term lines or tubes (6.5% versus 1.6) (Figure 2).
Regression analysis
Due to the bridle and non-bridle groups differing with regards to tube placement route, outcomes were re-tested for associations with bridle use. Nasal bridles reduced the risk of tube loss by 82% (OR [odds ratio] 0.18, 95% confidence interval [CI]: 0.09–0.035), improved tube, improved tube duration by 5.7 days (95% CI: 2.11–9.21) and tubes were three times more likely to reach redundancy when a bridle was used (OR: 3.0, 95% CI: 1.58–5.81).
Discussion
Main findings
When compared with tape securement, securing a tube with a nasal bridle from day 1 has been found to be independently associated with reduced tube loss, increased duration of tube use and the likelihood that the tube will reach redundancy and no longer be required. In a large retrospective observational study of ICU patients, inadvertent tube loss reduced from 53% to 9% when NI tubes were secured with a nasal bridle (Taylor et al, 2018). Monthly NG tube loss rates on ICU at Southmead Hospital declined by more than 50% within the first year after the use of nasal bridles was introduced (Taylor et al, 2018).
In the current audit, nasally sited feeding tubes secured by a bridle improved duration of tube use from 5 days to 9 days. In a similar population, Seder et al (2010) found that tubes secured with a nasal bridle had a median of three more feeding days compared with those secured with adhesive tape. Retaining tubes for longer could explain why nasal bridles improve the amount of energy and protein delivered to patients across the acute hospital setting (Beavan et al, 2010; Seder et al, 2010; Allan et al, 2018; Li et al, 2018). In a small retrospective service evaluation that measured nutritional delivery in ICU patients fed using NI tubes, patients were significantly more likely to meet protein targets if the tube was secured with a bridle compared with tubes secured without one (75% versus 55%) (Allan et al, 2018). Within an ICU population, Seder et al (2010) found that percentage calorie goal (how close patients get to meeting their target calorie intake) increased from 62% to 78% when a nasal bridle was used (P=0.016). It is important to note that tubes can still be inadvertently removed despite being secured with a bridle. This is usually as the result of a poorly fastened bridle or due to the patient pulling at the tube, narrowing the lumen and allowing the tube to slip through the clip.
Safety
The number of tubes used per patient was not significantly different between the two patient groups. This is likely to be due to the policy of securing the tube with a nasal bridle, if one tube has been removed. The audit described in this article found that the use of bridles significantly reduced inadvertent tube removal (57.1% unbridled versus 17.4% bridled). Similar studies reviewing tube loss also found that, when bridles were used, fewer tubes were inserted per patient (Beavan et al, 2010; Seder et al, 2010). The bridle should reduce tube placement risk, particularly when there has been previous misplacement. The risk of pneumothorax is 5.1% following one tube misplacement, but this increases to 36.4% after three or more tube misplacements (Marderstein et al, 2004).
Although not reported within the current audit, evidence from a meta-analysis, which included two small studies (Bechtold et al 2014), suggested that the rates of sinusitis were similar between patients with tubes secured with or without a bridle. However, this only included two small studies. No internal cartilaginous or soft-tissue injuries were identified by endoscopy in patients with bridles (Seder et al, 2010). The meta-analysis (Bechtold et al, 2014) did find that there was a possibility of increased risk of nasal bridles causing skin complications such as erythema and ulceration (13% versus 3% in patients without bridles), but the risk was reduced if correct insertion practices existed—to prevent the bridle being secured too tightly against the nares and to alleviate tension on the tape—and there was regular monitoring of the bridle. It is important to note that, while ensuring ICU patients are receiving adequate enteral nutrition, pressure ulcer risk (locally 0.11%) must be monitored when using a nasal bridle. Pressure ulcer data were not collected in the study reported here, but is reported within some larger RCT studies and meta-analyses (Bechtold, 2014; Brugnolli et al, 2014).
Cost
There was no cost-benefit analysis in the current study. However, previous research (Brazier et al, 2017) has estimated that compared with a mixture of tape securement and mittens and/or staff restraint in the stroke population, the pre-emptive use of nasal bridles may reduce enteral equipment treatment costs by 55%. This may underestimate cost savings from bridle-associated increases in feeding and medication treatment efficiency. In burns patients (n=41), nasal bridles have been shown to significantly reduce the hours of missed feeding (P=0.05) and the number of X-rays required to confirm tube position, P=0.003 (Li et al, 2018). In stroke patients, using mittens did not appear to prevent tube dislodgement (Brazier et al, 2017).
Limitations
In this study, the patient group numbers were small and restricted to a single centre. Consequently, the large increase in tube retention associated with pre-emptive bridle use warrants further study.
Conclusion
Nasal bridles appear to improve NG and NI tube retention in critically ill patients. Improved tube retention using a bridle is likely to reduce the number of tube misplacements (such as the tube entering the lung) because fewer tubes are needed per patient. Additionally, bridles improve nutrition and drug delivery and therefore treatment efficacy. As a result of the findings reported in this article, feeding tubes on ICU at Southmead Hospital are now pre-emptively secured using bridles on day 1 for all ICU patients fed through NG or NI tubes. Future studies are planned to assess whether this change in bridling practice has improved tube retention on ICU.