Skip to main content
Free AccessFull Paper

X-ray checks of NG tube position: a case for guided tube placement

Published Online:https://doi.org/10.1259/bjr.20210432

Abstract

Objectives:

Checking nasogastric (NG) tube position by X-ray is too late to prevent 1.5% of blind tube placements entering the lung and results in delays to feeding and drugs. We audit the safety of the tube position and delay incurred by X-ray.

Methods:

From Radiology reports, we determined whether tube position was safe for feeding, factors associated with an X-ray request and the time delay from X-ray request to that report. For tubes misplaced into the lung, the distance from the carina to tube tip was measured and compared with that from published records of guided tube placement.

Results:

From 1 July 2019 to 30 June 2020, 1934 X-rays were done to check NG tube position in 891 patients. Gastric placement was confirmed in 85% but, because of tube proximity to the oesophagus, only 73% were deemed safe to feed. The 2.2% of tubes reported to be in the lung were a median of 18 cm beyond the carina compared to 12 cm and 0 cm for electromagnetic and direct vision methods of guided placement. X-ray checks delayed feed and drug treatment by >2 h in 51% of placements and 33% of patients required >3 X-rays during their enteral episode.

Conclusion:

X-ray checks are common and detect a high percentage of unsafe tube placements, leading to repeated X-ray and delayed delivery of drugs and nutrition. Interpretation can be difficult even when following standard national criteria and post-placement X-ray cannot prevent deep lung placement. Guided or combined methods of confirming tube placement should be investigated.

Advances in knowledge:

Reports included 27.5% of placements as unsafe, 2.2% in the lung at a median depth of 18 cm beyond the carina and too late to prevent 7 pneumothoraces. X-rays were repeated >3 times in 33% of patients over their enteral course and we are associated with clinically significant delays to drug treatment (and nutrition) in 51%; combined methods of tube confirmation or guided placement may be safer and more efficient.

Introduction

A chest X-ray taken specifically to determine nasogastric (NG) tube position is considered the 'gold-standard' in terms of safety. To attain this safety, X-ray interpretation must only be undertaken by trained staff with adequate experience, using a tube that is radio-opaque along its entire length, where the tip position is captured on the X-ray and where four anatomical landmarks are identified.1 These landmarks are that the tube path: Follows the oesophagus, not bronchi; bisects the carina or bronchi; crosses the diaphragm at the sagittal midline and the tube tip is 10 cm below the left hemi-diaphragm.

Despite this, the first-line check for NG position in the United Kingdom (UK) is the pH of aspirated fluid, a surrogate for gastric tube position. Theoretically, gastric fluid may be regurgitated or vomited into the oesophagus or lung and then aspirated from a misplaced tube. In part for this reason, guidance issued during the COVID-19 pandemic suggested that an X-ray rather than pH be used to confirm initial tube position because of the perceived greater risk of regurgitation when a patient is turned or in a prone position.2 In addition, using a pH threshold of 5.5 carries a 12% risk that when the actual pH is 6.0 and potentially from the lung, it is interpreted as 5.5.3 Also, statistical modelling shows that a pH threshold of 5.5 carries a 50% risk of misidentifying oesophageal placement as being gastric and that lowering the 'gastric threshold' to 4.0 could reduce misidentification to 3.5% albeit with an increase in unnecessary X-ray use from 24 to 34%.4 A pH threshold of 5.0 has been suggested to minimise use of X-ray and oesophageal placement.5 This estimation was based on 15% of tubes being in the lung and 15% in the oesophagus. However, lung placement is shown at the end of procedure on X-ray in 1.5%6 and during procedure by guided tube placement in 10–35%,7,8 while end of procedure on X-ray shows oesophageal placement in 19–23%.9,10 An optimal pH threshold is therefore difficult to estimate.

Unfortunately, X-ray misinterpretation is the single largest (46.2%) cause of undetected lung misplacement.1,11 By comparison, pH was responsible in 21% of undetected cases, but only 12.3% when excluding cases where the pH of the aspirate was >5.5 or the tube had been previously flushed. Lastly, end-of-procedure X-ray cannot pre-empt complications that occur during placement and must be combined with other techniques to check for subsequent migration.12 In contrast, guided tube placement, using an electromagnetic trace of the tube path or direct vision, can discern tube position in real time, without delay, but are not sanctioned for use to confirm NG tube position in the UK.

From X-ray reports, we audited NG tube position, what factors might affect the number of X-rays required and the delays between X-ray request and confirmation of tube position. Depth of lung misplacement and delay to confirmation for X-ray were compared with that published for guided tube placement.8,13

Methods

Records

In a large tertiary care hospital, Radiology reports of NG tube checks were identified from the Computerised Radiology Information System (CRIS) between 1 July 2019 and 30 June 2020. From each report by a consultant Radiologist, the statement of the tube position and whether it was safe (gastric) was recorded by author ST. Where X-rayed tubes were misplaced within the lung the distance from the carina to tube tip was compared with previously published data of electromagnetically- (Cortrak) or visually guided (Kangaroo feeding tube with IRIS technology: 'IRIS') tube placement.8–13 This was done by marking the carina (AM: Consultant Intensivist) on X-rays or the bronchus deviation on the Cortrak trace (ST: Research Dietitian). The tube penetration beyond the carina was measured against the scale on the X-ray or Cortrak trace, respectively. For visually-guided (Kangaroo feeding tube with IRIS technology: 'IRIS') tube placement, the carina was clearly visible, therefore the tube was never placed beyond it (ST). We determined the delay from the X-ray request to the Radiologist’s report, what factors affect this delay and whether it exceeded the hospital’s threshold of 2 h for critical drugs.

Analysis

We tested continuous variables for normal distribution using the Shapiro-Wilk test (R Studio v. 1.1.463) and, because most were not normally distributed, we present data as median (interquartile range, IQR) and percentage. Continuous and categorical data were compared using two-tailed Wilcoxon rank-sum and Fisher exact tests, respectively. The effect of age, sex, ward or time of day on the number of X-rays per patient (an 'enteral episode') and the delay from X-ray request to confirmation of tube position was determined by linear regression and the likelihood of the tube being in an unsafe position, by logistic regression.

Ethics

As a registered UK audit (CA64043) of standard practice this study did not require ethics board approval.

Results

Over 365 days, including the first 2 month 'COVID-19 surge', 1934 X-rays were done in 891 patients (ICU: 54.3% vs ward: 45.7%) specifically to confirm NG tube position. ICU patients were significantly younger than ward patients (60.5y vs 74.4y, p < 0.0001; all patients: median 68.31y [IQR: 52.5–77.7]) but there was a similar male:female ratio (57.5:42.5%). Although most patients only required up to two X-rays, 33% required three or more, the maximum being 24, over their enteral episode. And, though more ICU patients required an X-ray, ward patients required more X-rays per person (ward: 2 [IQR: 1–4]; ICU: 1 [1-2], p < 0.0001). Age is associated with a requirement for more X-ray per enteral episode and being on ICU fewer X-rays (Table 1). It is noteworthy that ICU secure tubes with nasal bridles pre-emptively; wards rarely do.

Table 1. Regression models: factors affecting the number of X-rays required per patient, delay to confirming tube position and likelihood of unsafe tube position

VariableNumber of X-raysDelayUnsafe position
Effectp-valueEffectp-valueEffectp-value
Age0.0120.0001−0.030.36−0.010.002
Sex (male)0.090.40.080.94−0.270.035
ICU−1.47<0.0001−2.580.04−0.070.62
X-ray_daytime0.300.06−6.52<0.00010.480.004
Intercept2.05<0.000114.18<0.00012.21<0.0001
Variance explained11%-9%---

Most (84.5%) X-rays were done 08:00–20:00 but were disproportionately outside these hours on ICU (34.6% vs  3.6% ward) (Figure 1). Ward X-rays only peaked after 16:00 meaning that where tubes were misplaced, there was a risk safe tube placement would not occur before the 20:00 cut-off. The median delay from X-ray request to X-ray, report and confirmation of position was 0.17, 1.6 and 2.1 (IQR: 1.2–5.0) h, respectively. The time from request to confirmation was significantly negatively associated with being on ICU (−2.6 h) and the request being in the daytime (−6.5 h); age and sex were not significantly associated (Table 1). Delay from X-ray request to confirmation exceeded critical drug time thresholds of 2 h (systemic antibiotic, anti retroviral, chemotherapy) in 51% of cases.14 The above underestimates the problem of delay by not including the time from needing the tube to the request. In addition, in the 16.6% of X-rays where a follow-up X-ray was required because of tube misplacement, the median delay to confirmation increased to 4.8 h (IQR: 3–10.4).

Figure 1.
Figure 1.

Percentage of X-rays done over the time of day.

Most tubes were sited in the stomach (Table 2). However, after subtracting tubes that are less than 10 cm inside the stomach (n = 237) or coiled back towards the oesophagus (n = 10) only 72.5% were classed as safe. A total of 15.9% of tubes required advancing of which 76.9% were gastric, 13.3% at the GOJ, 0.3% hiatus hernia, 8.4% oesophageal, 0.3% pharyngeal and 0.6% unknown. This included gastric tubes < 10 cm from the GOJ, though 10 of these were classed as safe. Conversely, in five cases, the report advised removal where the tube was correctly placed or only needed advancing or removal of a kink; removal and replacement entails the greater risk of lung misplacement. Fewer tubes were coiled (7.5%), most of which had to be removed and repositioned because they were not gastric: 44.1% gastric, 6.9% GOJ, 2.8% hiatus hernia, 31% oesophageal, 1.4% mediastinum (upper), 9% pharyngeal, 0.7% lung and 4.1% unknown. 2% of tubes entered the pylorus or duodenum (Figure 2). In a regression model an unsafe tube position was negatively associated with age and being male but positively associated with daytime X-ray (Table 1).

Table 2. Tube position

PositionN%
Uncertain492.5
Pharynx150.8
Mediastinum50.3
Lung432.2
Oesophagus733.8
GOJ562.9
Hiatus hernia60.3
Stomach165085.3
Intestine382.0
Figure 2.
Figure 2.

X-ray showing safe and unsafe tube placement.

Of most concern, 43 (2.2%) tubes were misplaced in the lung. These 'blind' misplacements were a median of 18 cm (IQR: 16–23 cm) beyond the main carina, deeper than published records of guided tube placement using Cortrak (12 cm [IQR: 9–15], p < 0.0001) or IRIS (0 [IQR: 0–0], p 0.001) (Figure 3).8,13 New pneumothorax was detected immediately following tube placement in 7 (0.4%) and new pneumonia in 2 (0.1%) cases. In two cases of pneumothorax, end of procedure X-ray detected the tube in the oesophagus or stomach, possibly after it had been repositioned from the lung. All lung misplacements were detected before tube use; there were no 'NEVER' events associated with either X-ray or pH confirmation during this period. However, radiologists were uncertain of position in 2.5%, regardless of sex and despite using the 4 NHSI criteria to interpret NG tube X-rays.

Figure 3.
Figure 3.

Depth of tube misplacement in the lung using: A) Blind technique, (B) Cortrak, (C) IRIS.

Discussion

Main findings

Despite X-ray being the second-line check for gastric placement, it was used 1934 times in 891 patients over 365 days. Data were lacking in how many tubes were confirmed as gastric by aspiration of fluid with a pH <4.0 and how many X-rays were done having failed this test. However, previous studies in this institution indicate X-ray is used in ~50% of cases.15,16 Although most tubes were gastric (85.3%), only 72.5% were a safe distance into the stomach. Nevertheless the 8.1% of tubes that were in the GI tract but pre-gastric was lower than previous studies at this (17%) or other centres (23%).10,17 More tubes may now reach the stomach at this centre because the nose-ear-xiphisternum (NEX) method of estimating gastric tube distance has been replaced by xiphisternum-ear-nose (XEN) +10 cm.18 However, all nose to stomach tube lengths estimated using external body measurements could potentially result in oesophageal placement (NEX: 96.3%, Hanson-A: 99.5%, Hanson-B: 86.9%, XEN +10: 43.2%) and are therefore clinically unreliable.19 Indeed, the 8.1% failing to reach the stomach remains fourfold greater than the 2% over shooting into the intestine. Conversely, where tubes are curling anti clockwise in the stomach, insertion of longer lengths of tube will increase the proportion that will move towards the oesophagus.

The higher number of X-rays needed per non-ICU patient may reflect lower use of nasal bridles as studies have shown that pre-emptive use of bridles on ICU reduced inadvertent tube loss,20 and therefore subsequent X-ray use. The reason that the delay from request to X-ray confirmation is shorter on ICU and during the daytime is probably due to availability of medical, nursing and radiological staff. However, because most non-ICU X-rays occur from mid-afternoon it risks falling outside the 20:00 threshold for Radiologist cover and indicates a need for earlier tube placement to avoid long treatment delays. In contrast, ICU’s 24 h X-ray use could only be avoided if guided tube placement was available. Uncertain tube position in 2.5% of cases contributed to the need to re-X-ray because the tip was off-screen or the tube was not visible along its entire length. This occurred despite almost all tube walls containing 20–40% barium and radiologists interpreting X-rays by the 4 NHSI criteria. This shows that even under optimal conditions X-ray confirmation may fail. Previous study found that being male (length causing the tube to be off-screen), high BMI and guidewire removal, reduced tube visibility and radio-opacity and were associated with failure to confirm position by X-ray.21 The current audit found no difference due to being male but did not measure the other factors.

Lung misplacement remains common (2.2%), similar to previous studies.6,22,23 It also typical for a third of lung misplacements to result in pneumothorax and/ or pneumonia.6,22,23 In the current audit, the two cases of new pneumonia were not definitely attributable to prior lung misplacement. However, there was no formal follow-up to check for complications, so the seven pneumothoraces and two pneumonias may be an underestimate. The new pneumothorax detected when one tube was oesophageal and one gastric are examples of potential underestimate. This is where clinical symptoms during lung misplacement triggered repositioning of the tube but the subsequent X-ray showing a pneumothorax might not be linked to the initial misplacement because the tube was now correctly positioned. These complications are unsurprising as misplaced tubes were a median of 18 cm beyond the carina, enough distance to cause trauma or take nasopharyngeal bacteria deep into the lung. This is significantly deeper than when using real-time placement, particularly direct vision.7,8 When pH or X-ray are used to check final tube position, deep lung misplacement can only be avoided if combined with CO2 detection at a tube depth of 30 cm24 or X-ray at 40 cm, post-carina.25 This is important because pneumothorax is associated with deep lung penetration.26 However, although CO2 detection significantly reduced deep lung placement, it can still fail to detect CO2 or fail to warn in time to prevent complications.20 In addition, carina depth varies by 8.5 cm in adults,27 so the tube length required to show deviation into a bronchus on X-ray is uncertain; repeat X-rays would incur further irradiation and delays. And, neither method helps to actively guide the tube to avoid leaving it in the oesophagus. It is unclear why the tube being in an unsafe position was negatively associated with age and being male and positively associated with a daytime X-ray; this should be re-evaluated in a larger study.

As with X-ray, electromagnetic-guidance cannot detect lung placement until there is deviation from the sagittal midline. When tube deviation is subtle, it may only be interpreted as a misplacement when it is deep.8 In contrast, misdirection of IRIS tubes into the respiratory tract was always detected pre-carina, within the trachea.13 Potentially critical drugs and feed were delayed more than 2 h for 51% of individual X-rays. But, in 16.6% cases where lung or pre-gastric placement led to tube removal and replacement, further X-rays and delays increased the median delay to 4.8 h. Because 33% of patients require >3 X-rays, during an enteral episode, the cumulative loss of drug and nutritional treatment can be significant; in stroke patients, 18% of feeding time was lost.28 Guided placement, by contrast, confirms tube position at placement, without delay.8 In addition, though chest X-rays per se are relatively safe they present a significant risk from clinical events when done off-ward.29 The above problems are likely underestimated because some patients will have had X-rays checks prior to and after the 1 year audit period.

X-ray use on ICU was spread over the 24 h, including overnight, because NG tube access is seen as vital. Ward X-rays peak after 16:00 and indicate that an earlier request would facilitate confirmation when radiologists are available up to 20:00.

Data collection covered the early COVID-19 pandemic up to 30 June 2020. COVID patients more often present with lung opacity, necessitating higher radiation to adequately visualise tubes.30 The infection risk may mitigate against X-ray use. There is conjecture that prone positioning may increase aspiration risk and falsely indicate gastric position when using a pH check.2 However, prone-positioning per se is not associated with increased gastric residual volumes (GRVs) or delayed gastric emptying.31–33

Limitations

This audit was limited to 1 year, in a single centre and we could not ascertain the number of NG tubes confirmed by pH. However, high X-ray use means that estimates of tube position and the delayed reporting are likely to be typical for this procedure.

Conclusion

Despite X-ray being the second-line check for gastric placement, X-ray use was high and detected 27.5% of placements as unsafe. 33% of patients required >3 X-rays during their enteral episode. Drug and nutrition treatment was delayed >2 h in 51% of X-rays. When tube misplacement necessitated a repeat X-ray, the median delay was 4.8 h and represented a high work burden to medical, nursing and radiological staff. Tube position was uncertain in 2.5% of X-rays, showing that even expert, NHSI-driven X-ray interpretation can fail. Improved safety, timeliness and cost may result by combining, as staffing permits, expert-led, guided-tube placement or CO2 detection at 30 cm followed by pH or X-ray checks at the end of procedure.

Acknowledgment Our thanks to Gillian Forte, Glyn Short and Janet Iles for extracting the dataset.

Authorship contributions Conception and design of the study, the acquisition of data (SJT), interpretation of X-rays (ARM) and interpretation of the data (both). Drafted or provided critical revision of the article (both). Provided final approval of the version submitted for publication (both).

Conflict of interest ST did consultancy for Cortrak in 2008, worked on a studies in 2013-14 sponsored by Cortrak and 2020-current sponsored by Cardinal Healthcare, both through North Bristol NHS Trust. Cortrak and Cardinal played no part in sponsoring, planning, execution or publication of the current work. ARM: None.

REFERENCES

Volume 94, Issue 1124August 2021

© 2021 The Authors. Published by the British Institute of Radiology


History

  • ReceivedApril 06,2021
  • RevisedMay 12,2021
  • AcceptedMay 17,2021
  • Published onlineJuly 08,2021

Metrics