Impact of introducing transnasal endoscopy on expanding diagnostic endoscopy services

Abstract

Introduction

Demand for endoscopic services is outstripping capacity in the UK. Transnasal endoscopy (TNE) utilises a narrow calibre endoscope to pass through the nasal passages, thereby reducing retching and discomfort. It is better tolerated compared with standard transoral endoscopy (TOE) but is still rarely used as a diagnostic modality. There is still uncertainty about how well it performs against TOE in diagnostic ability, cost and efficiency.

Methods

We explored utilising TNE to deal with the growing demand for endoscopy. We compared findings to TOE procedures undertaken during the same time period. We evaluated cost differences, duration/time in the department for procedures as well as quality of procedures (both in terms of performance indicators as well as image quality using the validated POLPREP score).

Results

A total of 241 upper gastrointestinal endoscopy procedures were evaluated (100 TNE and 141 matched TOE) between December 2021 and February 2022. TNE outperformed TOE in obtaining >95% success rate in nationally approved key performance indicators (retroflexion and duodenal intubation, both p=0.026). It also was associated with better image quality in the oesophagus with a POLPREP A3 Score (excellent image quality) in 36/98 available images compared with 26/136 TOE (p=0.028) and was equivalent to TOE in the stomach and duodenum. TNE was identified as having a key role in facilitating complex cases, previously failed via the TOE route with a success rate in 11/12 (91.7%) of such cases. TNE also shows promise in correcting misdiagnoses of short-segment Barrett’s oesophagus (39%) compared with TOE (14.8%) (p=0.087).

Conclusion

TNE is an emerging endoscopic modality, which shows great promise in replacing TOE in most diagnostic circumstances. In a modern healthcare service, TNE is cheaper, better tolerated and outperforms TOE in multiple domains.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Transnasal endoscopy (TNE) has been demonstrated to be a safe, effective and well-tolerated alternative to conventional transoral endoscopy. However, despite this TNE remains underutilised within the UK.

WHAT THIS STUDY ADDS

• This paper further explores the viability of introducing a TNE service. Assessment of image quality, time comparison for procedures, workforce demand and costs have not been previously published in relation to TNE.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• TNE could be utilised to relieve mounting pressures on endoscopy services and aid general throughput of patients. It could be used as an adjunct to traditional endoscopy to meet significant demands or used as first-line investigation to expand existing endoscopic services. It may also be able to facilitate the identification of individuals with previously misdiagnosed Barrett’s oesophagus, thereby further reducing demands for surveillance endoscopy.

Introduction

Diagnostic endoscopy services

Gastrointestinal endoscopic procedures are the first-line tool for detecting GI pathology.^{1 2} Conventional transoral endoscopy (TOE) is the current standard diagnostic method used to visualise the oesophagus, stomach and duodenum. It can be performed without sedation, using local anaesthesia and/or with sedation for increased tolerability and comfort. Sedated TOE is used more frequently in developed countries. The vast majority of endoscopists report using sedation in most procedures.3,5 While increasing tolerability and quality of examination, the use of sedation increases the risk of adverse cardiorespiratory events,⁶ especially in frail patients with multiple comorbidities. Patients receiving sedation require careful patient monitoring, which increases the time of admission, cost, and the number of staff required for a procedure.47,11

Increasing demand for endoscopy

There is substantial demand for diagnostic endoscopic procedures within the UK, which preceded the COVID-19 pandemic and was exacerbated afterwards. In the first 3 months of 2021, only 57.9% of NHS services met urgent cancer waits, 17.9% met routine waits and 13.4% met surveillance targets.^{12 13} Demand is outstripping capacity, with current modelling suggesting that this demand is likely to increase further.²

Transnasal endoscopy as an alternative diagnostic approach

Transnasal endoscopy (TNE) has been introduced as an alternative method to diagnostic TOE. It is performed using an ultrathin endoscope (<6 mm diameter) inserted via the nasal orifice. This minimises stimulation of the gag reflex, resulting in less sympathetic stimulation and cardiovascular stress.¹⁴ This increased tolerability has been shown to allow for cheaper and more time-efficient procedures, by avoiding sedation.15,17 The absence of sedation allows patients to be discharged immediately following endoscopy and for TNE to be performed in a wider variety of settings, with no specific endoscopy suite or theatre required.^{17 18}

Technical advancements in visual resolution have allowed TNE to improve on previous concerns regarding optic resolution. The recently published British Society of Gastroenterology Sedation Guidelines¹⁹ have endorsed the use of TNE as an acceptable alternative to TOE for routine diagnostic endoscopy, but despite this, TNE remains rarely used, comprising only 3% of all endoscopic procedures in the UK.^{2 18}

Study intentions

The primary purpose of this study was to explore the effectiveness of a TNE service in daily practice, evaluate the potential of TNE in alleviating the mounting pressure on diagnostic endoscopy services, assess its technical performance against standard TOE and better understand any disease-specific indications for use.

Methods

General methods/service setup

To establish viability, a TNE pilot programme took place at University Hospital Llandough, a Cardiff University Teaching Hospital with four-endoscopy rooms. Funding was provided by the Moondance Cancer Initiative²⁰ and transnasal scopes were provided by Olympus UK and Ireland. The pilot was operational between December 2021 and February 2022, involving four members of staff and approximately 1–2 endoscopy lists a week. Lists were run with a single endoscopist supported by one trained nurse.

Patients were identified for upper gastrointestinal endoscopy via the TNE route from the waiting lists, using previously defined inclusion and exclusion criteria.¹⁵, ²¹

Patient pathway

Patients underwent their procedure in a capsule endoscopy room on days when this clinical space was not in use, thereby increasing the capacity for endoscopy by adding an additional room for activity on those days. After booking into reception at the endoscopy suite, patients were consented for their procedure, and given a gastric cleaning drink—‘The Cardiff mix’, which is standard practice in our endoscopy unit to act as a mucolytic and improve luminal views. There is data on the use of combination of N-acetylcysteine with simeticone as a mucolytic, however the constitution of this mixture has not been explored. For our pilot, we used a mixture made from 1 L of water, 30 drops of infacol (Simeticone), 40 mL of N-acetylcysteine (200 mg/mL). Patients were given 50 mL to drink prior to endoscopy. Patients were then treated with lidocaine/phenylephrine nasal spray as per standard technique¹⁷ approximately 10 min before the procedure.

A standard upper GI endoscopy was performed detailing all conventional endoscopic landmarks as per the indication of the referral using a 5.4 mm diameter Olympus GIF-H190N transnasal gastroscopes and an Olympus EVIS X1 processor. After the procedure, a tablet with a validated questionnaire^{22 23} was completed by patients, while the endoscopist produced a report.

Comparison group

As part of the analysis, these procedures were compared with consecutive TOE procedures in the same time frame performed by the same endoscopy team using Olympus GIF-H290Z 10.2 mm gastroscopes. Comparison of TNE and TOE scope characteristics are highlighted in online supplemental table 2.

Data collection and analysis

In addition to standard demographic information and referral reason/outcome, data were collected on three key areas: patient satisfaction, endoscopy key performance indices (KPIs) and endoscopy image quality. Patient satisfaction was assessed via the completion of an adapted Patient Reported Outcome Measures (PROMS)-validated GI endoscopy satisfaction questionnaire post-procedure, the Gastrointestinal Endoscopy Satisfaction Questionnaire (GESQ) patient-comfort scoring system.^{22 23} Endoscopy KPIs of each procedure were collected using the endoscopy reporting system Clinical Workflow Manager (CWM), Fujifilm operator productivity and endoscopy unit data reports to measure performance against the National Joint Advisory Group on GI Endoscopy (JAG) KPIs (data points: oesophageal intubation, J manoeuvre rate (retroflexion), D2 intubation).²⁴ The rate of cervical inlet patch detection was also documented as a novel KPI. The quality of endoscopic images was independently scored by two authors using the POLPREP scale, a validated scoring system assessing mucosal visibility of the upper gastrointestinal tract.²⁰ Image quality was scored based on the best possible images captured/saved.

Data were analysed using the Statistical Package for the Social Sciences, SPSS (IBM) V.22.0. Categorical variables were analysed and compared using χ² tests. For continuous data, normality was assessed using the Kolmogorov-Smirnov test and parametric data was analysed using a paired t-test. Non-parametric data were analysed using a Mann-Whitney U test. Statistical significance was set at p<0.05.

Ethics statement

All patients had capacity consented for TNE and for their anonymised data to be collected and utilised within the following publication. Patients were not involved in the initial design of the pilot programme, but patients’ feedback was actively used in optimising the endoscopy procedures during the study.

Results

Demographics

Between December 2021 and February 2022, 120 patients were invited to participate in the pilot. 101 patients attended for TNE; 1 patient withdrew consent in the waiting room, resulting in 100 patients undergoing the procedure. The median age of patients was 64 years (range 24–93). This included 57 male patients and 43 female patients.

During the same period, 141 diagnostic TOE cases were also performed by the same cohort of endoscopists. The median age of the comparator TOE group was 61 (range 18–92). This included 70 male and 71 female patients. There was no statistically significant difference between age and sex of patients (p=0.673 and p=0.260, respectively) between the two cohorts.

Procedure indication

TNE and TOE were performed for a variety of indications on four main pathways (urgent suspected cancer (USC, 2-week wait), urgent, routine and surveillance). Of the TNE cohort (n=100), 10 patients were scoped under the USC pathway, 14 on the urgent pathway, 10 in the routine and 66 on the surveillance pathway. In contrast to the TOE group (n=141), 77 patients were scoped under the USC pathway, 27 on the urgent pathway, 4 in the routine and 33 on the surveillance pathway.

Of the 66 surveillance cases performed with TNE, 41 were for Barrett’s surveillance, 10 for follow-up of gastric polyp/neuroendocrine tumours, 8 for gastric mapping/assessment of atrophy/intestinal metasplasia, 4 for oesophagitis/ulcer follow-up, 2 for eosinophilic oesophagitis repeat biopsies and 1 MALT lymphoma follow-up. Figure 1 provides further breakdown of the USC, urgent and routine cases.

Figure 1. Breakdown of non-surveillance cases in pilot. NET, Neuroendocrine tumour; GORD, Gastrointestinal reflux disease.

TOE was performed for a variety of indications during the pilot. The breakdown of indications for both TOE and TNE is shown within table 1.

Table 1. Summary of indications for TOE (n= 141) and TNE (n=100) cohorts.

Indication	TNE (n=100)	TOE (n=141)	Total procedures
Barrett’s surveillance
Count:	41	27	68
% within indication:	60.30%	39.70%
% within procedure:	41.00%	19.10%
Gastric mapping/assessment
Count:	8	3	11
% within indication:	72.70%	27.30%
% within procedure:	8.00%	2.10%
Grade C/D oesophagitis/gastric ulcer
Follow-up
Count:	8	12	20
% within indication:	40.00%	60.00%
% within procedure:	8.00%	8.50%
Polyp/gastric NET follow-up
Count:	11	1	12
% within indication:	91.70%	8.30%
% within procedure:	11.00%	0.70%
Dysphagia
Count:	5	30	35
% within indication:	14.30%	85.70%
% within procedure:	5.00%	21.30%
Eosinophilic oesophagitis follow-up
Count:	2	2	4
% within indication:	50.00%	50.00%
% within procedure:	2.00%	1.40%
Dyspepia/abdominal pain/weight loss
Count:	11	32	43
% within indication:	25.60%	74.40%
% within procedure:	11.00%	22.70%
Abnormal radiology
Count:	3	4	7
% within indication:	42.90%	57.10%
% within procedure:	3.00%	2.80%
Coeliac disease follow-up
Count:	2	5	7
% within indication:	28.60%	71.40%
% within procedure:	2.00%	3.50%
Nausea/vomiting/reflux
Count:	8	9	17
% within indication:	47.10%	52.90%
% within procedure:	8.00%	6.40%
MALT lymphoma follow-up
Count:	1	0	1
% within indication:	100.00%	0.00%
% within procedure:	1.00%	0.00%
Anaemia/?GI bleed
Count:	0	16	16
% within indication:	0.00%	100.00%
% within procedure:	0.00%	11.30%
Total	100	141	241

GIGastro-intestinalMALTMucosa-Associated Lymphoid TissueNETNeuroendocrine tumourTNEtransnasal endoscopyTOEtransoral endoscopy

There was a significant statistical difference (p<0.001) between the pathway priority and indication in which the TNE and TOE cases were performed. There was a greater number of surveillance cases in the TOE cases, with a large cohort of Barrett’s oesophagus patients undergoing TOE (Barrett’s oesophagus >3 cm excluded from TNE as per predefined exclusion criteria).¹⁵ TNE was not utilised in the exploration of potential UGI bleeds or cases of unexplained anaemia, owing to the potential need for endoscopic therapy via the TOE route.

Endoscopist characteristics and training

Of the 100 TNE procedures, 46 were listed to be performed by a TNE proficient consultant, 35 with such a consultant supervising a trainee in TNE (defined as <10 procedural experience) and 24 with a newly TNE trained practitioner (defined as >10 completed supervised cases). All endoscopists trained in TNE were experienced at TOE (>300 cases performed with full JAG certification for independent practice). The TOE comparison group was performed by the same cohort of endoscopists.

Procedural outcomes

Of the 100 TNE patients, 99 patients completed the endoscopy procedure (98 via TNE, with 1 case failing a nasal insertion due to challenging anatomy converting immediately to a TOE). One other patient withdrew consent during the procedure. This patient declined all attempts at endoscopy including TOE and therefore was excluded from further analysis (ie, 1% failure rate).

In the comparator TOE group, there were six failed procedures (4.3%), but this was not statistically more common than in TNE (p=0.138).

Endoscopic diagnosis

Of the 99 completed TNE procedures, the majority either revealed no abnormalities (normal) (35/99) or Barrett’s oesophagus (30/99). Gastric atrophy/IM (7/99), gastric ulcers (4/99), reflux oesophagitis (3/99), hiatus hernia (6/99), gastric polyps (3/99), gastritis/duodenitis (7/99) and eosinophilic oesophagitis (1/99) were also identified as the principal diagnosis following completed TNE. A sinister diagnosis was detected in 3/99 procedures (1 Barrett’s HGD, 1 gastric mass, 1 Gastric MALT lymphoma recurrence).

Biopsies were taken in 71 procedures of the 99 completed procedures, including 152 individual biopsy samples. All sets of biopsies were deemed sufficient for analysis and diagnosis by the reporting pathologist.

Barrett’s oesophagus

Of the 41 patients attending Barrett’s oesophagus surveillance under TNE, 16/41 (39.0%) patients were able to be discharged, 23/41 (56.1%) required repeat endoscopy (further surveillance), 1 patient was found to have a gastric mass requiring MDT input and clinic follow-up. 1 patient was found to have high-grade dysplasia Barrett’s oesophagus on TNE suitable for endoscopic treatment. This discharge rate compared with only 4/27 (14.8%) in TOE cohort, p=0.087, (figure 2A). The patients discharged from further Barrett’s surveillance were typically due to the diagnosis of an Irregular Z Line as opposed to true Barrett’s oesophagus. This was made possible by less retching with our TNE cohort in comparison to procedures performed by TOE (figure 2B).

Figure 2. (A) Clinical outcome of Barrett’s oesophagus surveillance via TNE (n=41) and TOE (n=27) during the pilot. (B) Example endoscopy images of same patient. TNE highlighted patients with small segment Barrett’s previously seen on TOE (ie, columnar mucosa in red above the top of the gastric folds dotted line) may have in fact been an irregular Z line enhanced by retching rather than true Barrett’s oesophagus. Image 1 representing TOE and image 2 representing TNE. TNE, transnasal endoscopy; TOE, transoral endoscopy.

‘Complex’ endoscopy

During the pilot period, 12 patients who had either previously failed an endoscopy (n=7) or were considered high risk for TOE due to existing comorbidity (n=5, a patient with severe agoraphobia and germaphobia, a case of complex duodenal structuring non-traversable with a standard 10.2 mm diameter gastroscope, a patient with complex neck anatomy (Klippel Feil syndrome), a case of cerebral palsy and a case of severe obesity (Body Mass Index, BMI >40)) were offered a TNE procedure.

Median number of previous attempts at endoscopy was 2 (range 1–3). In total, 11/12 (91.7%) of cases were successful using TNE. One failed procedure was due to anxiety in a patient previously unsuccessful on two occasions (subsequently referred for general anaesthesia/propofol endoscopy).

Quality of procedure

The quality of TNE was assessed using JAG criteria.²⁴ Minimum standards for oesophageal intubation, retroflexion/J manoeuvre and duodenal intubation were exceeded with 99/100 (99.0%) cases achieving success in all three quality standards (95% JAG target in all domains). This compared with 135/141 (95.7%) of cases achieving oesophageal intubation and 131/141 (92.9%) achieving retroflexion and D2 intubation in the TOE cohort.

These values represent significant differences in retroflexion (p=0.026) and D2 intubation (p=0.026) in favour of TNE. This perhaps was due to the greater volume of assessments for dysphagia within the TOE cohort but could feasibly be due to the ultra-thin scope of TNE allowing an increased ability to transverse strictures or tumours.

Furthermore, inlet patch detection was documented in 6/100 (6.0%) of TNE cases, compared with 1/141 (0.7%) of TOE cases. This significantly greater rate of inlet patch detection rate in TNE (p=0.015) was perhaps due to the increased comfort allowing for greater inspection of the region.

Appropriate images for analysis were captured/available for 98 of the TNE procedures and 136 of the TOE procedures and were independently assessed for quality using the POLPREP scale. ²⁰Images were scored separately based on anatomical location with oesophageal gastric and duodenal image assessments. The comparison of quality between these images and images under TOE for the same cohort is summarised in table 2.

Table 2. Oesophageal, gastric and duodenal image quality of TNE as compared to TOE using the POLPREP scale²⁰.

Anatomical location	Image quality of TNE (number of cases/98 (%))	Image quality of TOE (number of cases/136 (%))	P value
Oesophageal	A0-0 (0.0%)	A0-1 (0.7%)	0.028
	A1-4 (4.1%)	A1-10 (7.4%)
	A2-58 (59.2%)	A2-99 (72.8%)
	A3-36 (36.7%)	A3-26 (19.1%)
Gastric	B0-2 (2.0%)	B0-3 (2.4%)	0.218
	B1-16 (16.3%)	B1-9 (7.2%)
	B2-68 (69.4%)	B2-98 (78.4%)
	B3-12 (12.2%)	B3-15 (12.0%)
Duodenal	C0-0 (0.0%)	C0-1 (0.8%)	0.166
	C1-2 (2.0%)	C1-6 (4.9%)
	C2-52 (53.1%)	C2- 76 (61.8%)
	C3-44 (44.9%)	C3-40 (32.5%)

TNEtransnasal endoscopyTOEtransoral endoscopy

This comparison shows a statistically significant improvement in image quality of the oesophagus but no difference between images of the stomach or duodenum. This feasibly could have been due to the mucolytic drink given as standard within the TNE protocol.

Patient comfort

For the first time, the validated GESQ endoscopy comfort score was analysed in patients undergoing a TNE procedure. Pain described during the procedure (question answered by 95 respondents) was mild in 36/95 cases, moderate in 11/95 cases and severe in 2/95 cases, very severe in 0/94 cases with 46 patients describing no discomfort at all. Interrogation of causes for discomfort revealed 55/95 of patients reporting no retching, 37/95 described a little retching and 3/95 of those scoped described a lot of retching.

Ninety-four respondents completed a section of the questionnaire on general satisfaction, with 87.2% (82/94) satisfied with their TNE. Of these, 74/94 had previously undergone a TOE, and when asked if a further procedure was required, 70/74 (94.6%) stated they would prefer TNE over traditional TOE.

Duration

We wanted to evaluate the efficiency of TNE compared with TOE and assess the mean time spent in the department. Of all 98 completed TNE procedures, this was 59 min 20 s (24 min 35 s—2 hours 11 min 0 s). This involved a mean time of 47 min 19 s (10 min 11 s—1 hour 15 min 12 s) preprocedure and 12 min 3 s (4 min 22 s—44 min 6 s) in which the procedure was completed. All patients deemed the length of time to perform the endoscopy as acceptable.

This is compared with the mean TOE patient flow time of 55 min and 15 s (table 3) with longer times needed for TOE procedures undergoing detailed surveillance or propofol deep sedation procedures (a separate comparator group). Comparable times were seen between simple diagnostic TOE and TNE procedures.

Table 3. Mean time comparison of TOE and TNE in hours:min:s.

	TNE	TOE	Surveillance	Propofol
Pre-procedure	00:47:19	00:43:31	00:43:20	00:43:28
Procedure	00:12:03	00:10:57	00:13:53	00:11:51
Total time	00:59:20	00:55:15	00:56:19	02:25:03

TNE—all transnasal endoscopy procedures whatever the indication, TOE—transoral endoscopy for all urgent/suspected cancer cases, surveillance—routine transoral endoscopy repeat procedures for pre-cancerous investigation (Barrett’s oesophagus surveillance and gastric atrophy/mapping assessments), Propofol—transoral cases requiring anaesthetic support (via propofol deep sedation).

Cost

TNE was found to be significantly cheaper than traditional methods of endoscopy with an average cost per procedure of £106.43. This is compared with £151.35 with TOE. The cost disparity was found to be largely found due to the decreased number of nursing staff required with TNE; being performed with one band five nurses present as compared with 2 with TOE. See online supplemental table 1 for the full cost breakdown of workforce and consumable costs.

Complications

Out of the 100 TNE procedures, 2 were unable to be completed. One patient withdrew consent during the procedure refusing all further endoscopy (including TOE). In one patient, nasal intubation was unable to be achieved. In this case, a conversion to oral endoscopy was undertaken and successfully completed.

Three patients experienced physical complications of TNE: One patient experienced an episode of self-limiting epistaxis (1%), One patient experienced an episode of self-resolving chest pain (1%) and one patient had a vasovagal episode (1%). Despite these complications, all these patients fully completed the procedure and were subsequently discharged from the department.

Discussion

Despite previous data, TNE represents only 3% of all OGDs performed in the UK.^{2 18} The pilot programme suggests that TNE is a safe and effective method of expanding services to meet the current demand for endoscopic procedures.

The comfort of TNE has been well established as being the main reason for its use with minimal gagging or abdominal discomfort traditionally reported compared with conventional unsedated TOE. While other comfort assessments have been undertaken in the past, here, for the first time, we used the validated endoscopy comfort survey (GESQ) to assess patients undergoing TNE. General comfort scores were better in the TNE group with the majority of patients (94.6%) who had previously undergone both procedures stating a preference for TNE over TOE. This is consistent with previous data, highlighting the increased tolerability of TNE^{15 17} and further supporting its routine use as per patient preference.²⁵ Furthermore, when evaluating complication rates, the epistaxis rate seen in our TNE pilot was small and consistent with existing literature.^{18 26 27}

The other complication often quoted is failure to pass the scope through the nose (with a conversion to TOE required of 5%).¹⁸ In our experience of 100 patients, this was only 1%, demonstrating that in experienced hands, the failure rate is low. Interestingly there are also a group of patients that are challenging to manage due to comorbidity, anxiety or general difficulty in tolerating a TOE, even with the use of sedation medications such as a benzodiazepine like midazolam with/without the additional use of a short-acting opioid like fentanyl. This group of patients are at risk of delays in diagnosis and treatment and can often require multiple, potentially costly, attempts at endoscopy, or requirement for a general anaesthetic to facilitate a procedure, which again can be complex both in its organisation as well as it is potential for risk. We demonstrate the utility of TNE in this small, but clinically significant group of patients showing excellent success rates, suggesting this should be considered as a first-line investigation in such cases.

Existing literature has highlighted that TNE has comparable diagnostic potential as compared with TOE.^{15 17 18} In this pilot study, TNE outperformed TOE on JAG quality metrics. Furthermore, inlet patch detection (a novel surrogate marker for endoscopic quality and potential lesion detection) was documented in 6.0% of cases compared with 0.7% of TOE. This congenital endoscopic phenomenon is detectable in up to 10% of individuals,²⁸ and our data suggest that TNE allows for a better degree of mucosal inspection and affirms that a more comfortable patient could provide better diagnostic information.

Here, we also demonstrate the quality of images obtained using a smaller diameter scope combined with a mucolytic drink preprocedure provided comparable quality to conventional transoral endoscopy using the validated POLPREP score. This could allow for the flexibility of TNE to be used as part of the high-quality assessment of patients on waiting lists outside of the standard endoscopy unit footprint.

We acknowledge that due to limitations of resources, only Olympus scopes were utilised within this study. We believe that other scopes are likely to demonstrate similar quality resolution, but this would have to be studied further in future research.

In this study, TNE had significant advantages over TOE, as measured by a reduction in the workforce needed (as there is no need for recovery nursing to take place as patients are discharged from the procedure room immediately after their TNE) and a consequent reduction in costs. There are some other potential cost savings incurred with a reduction in medication and plastic use. This may also have additional benefits in reducing environmental impact. The time required for a TNE procedure appears to be no different from a standard TOE and is certainly more efficient than a general anaesthetic or sedated procedure.

We acknowledge that this pilot was undertaken during the COVID-19 pandemic. As such full PPE compliance was required. We hypothesise that implementing TNE without these constraints would have increased the speed and efficiency of performing procedures. However, more research is required to fully detail the efficiency of TNE outside of the COVID-19 pandemic.

In this study, we also demonstrated the potential value of TNE in correcting diagnoses of ultrashort segment (<1 cm) Barrett’s oesophagus. Of those undergoing Barrett’s surveillance, 39% of patients were able to be discharged with TNE as opposed to traditional surveillance of Barrett’s oesophagus, which would recommend regular 3–5 yearly endoscopy for this cohort of patients. This discharge rate is likely contributed to the decreased retching in TNE causing increased accuracy of diagnoses. On previous TOE, these patients may have had an irregular Z line rather than true Barrett’s oesophagus (columnar metaplasia) and the diagnosis of Barrett’s oesophagus may therefore have been overcalled on initial TOE (figure 2B). This increased discharge rate (and therefore reduced requirement for repeat surveillance procedures) may also help reduce the general carbon footprint of endoscopy, as the greenest endoscopy procedure is the one that you do not need to perform. Providing a sustainable endoscopy service has been identified as a key priority by the British Society of Gastroenterology.²⁹

We acknowledge some limitations in our study, namely that we were unable to fully match TNE procedures with TOE procedures. To maintain consistency of data, we felt it best to compare TOE procedures during the same timeframe and performed by the same endoscopists as this reduces the variation of endoscopy technique and biopsy protocol, which could affect the time of the procedure and impact comfort scores. As a result of this, we acknowledge also that the reason for referral was not matched between the two groups, with TNE procedures being performed in stable long waiting cases (typically for surveillance), TOE procedures being performed in more urgent cases, and those more likely to require therapeutic intervention. We also acknowledge that this study was a pilot study, while showing positive results further evaluation of TNE is essential to ensure that these results persist through all populations to provide sustained improvement.

Conclusions

We believe the data generated from this pilot study demonstrate the benefits of TNE over TOE, namely that it is a preferred, excellently tolerated, more cost-effective modality with comparable visualisation to TOE, while also having clear benefits in both facilitating early discharge of patients misdiagnosed with Barrett’s oesophagus as well as complex cases that may otherwise require additional resource to facilitate an anaesthetic-assisted procedure.

The data now support the utilisation of TNE to relieve mounting pressures on endoscopy services; being used as an adjunct to traditional endoscopy to meet significant demands or used as first-line investigation to expand existing endoscopic services.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.