Introduction
Oesophageal soft food bolus obstruction is a well-recognised and not uncommon presentation to emergency departments. While many obstructions are found to resolve spontaneously, there are cases that require surgical intervention.Reference Scott-Brown and Gleeson1 Swift management to resolve the obstruction is important, given the potential risk of rare but serious consequences such as perforation and mediastinitis.Reference Loh, Tan, Smith, Yeoh and Dong2
In general, initial hospital management can consist of an observation period with simultaneous medical therapy postulated to relax the oesophageal tone to permit the bolus to migrate distally.Reference Leopard, Fishpool and Winter3 Failing this, the still often practiced approach is rigid oesophagoscopy performed by ENT surgeons under general anaesthetic.
Access to and quality of flexible oesophagoscopy and/or oesophago-gastro duodenoscopy equipment has vastly improved. Given the advantage of negating the need for general anaesthesia, this procedure offers patients a quicker recovery with a smaller risk of complications.Reference Gmeiner, von Rahden, Meco, Hutter, Oberascher and Stein4 It is performed by general surgeons, gastroenterologists or nurse endoscopists.
The aim of this study was to review the current performance, pathways and outcomes of patients attending emergency departments across three health boards in Scotland to provide evidence-based practice in a condition that otherwise has a poor, historical or conflicted evidence base.
Materials and methods
This study was approved as an audit by the Research Ethical Committee and local Caldicott guardian approvals were gained. Data were requested from clinical informatics on all patients admitted with a soft food bolus obstruction in three Scottish territorial health boards, Greater Glasgow & Clyde, Lothian and Lanarkshire, for periods between January 2016 and December 2018. The International Classification of Diseases, Tenth Revision codes used were T18.1 (foreign body in oesophagus) and K22.2 (oesophageal obstruction). A retrospective case note review was carried out to record details of presentation, emergency department treatment, specialty management and final outcome for each patient, including any complications.
Results and analysis
We identified 377 patients admitted with suspected food bolus. Of these, 64 patients had a sharp and/or solid material obstruction and so were excluded. The mean age of the patients was 55 years (range, 7–93 years) and 63 per cent of patients were male (Table 1).
Patient characteristics included in this study
Overall, 35 per cent of patients reported prior history of food bolus obstruction, 12 per cent had a known oesophageal stricture and 8 per cent had a history of oesophageal malignancy. The most common medical co-morbidity was arterial hypertension, with 14 per cent of patients having this diagnosis. In addition, 8 per cent of patients had ischaemic heart disease.
Ultimately, approximately two-thirds of patients admitted required therapeutic oesophago-gastro duodenoscopy, 18 per cent had a rigid oesophagoscopy and 16 per cent had spontaneous resolution (Table 2). In the spontaneous resolution group, 28 per cent of patients received no medical management and 72 per cent received Buscopan. Data on pre-existing arterial hypertension and ischaemic heart disease were collected as these are contraindications to Buscopan. Despite this spread of final intervention, 37 per cent of patients were initially admitted under ENT, 35 per cent to general surgery and 28 per cent to gastroenterology (Figure 1).
Final outcome of treatment
Initial and subsequent referrals. OGD = oesophagogastro duodenoscopy
Of those initially admitted under ENT, 23 per cent had spontaneous resolution, 47 per cent underwent rigid oesophagoscopy and a further 30 per cent were referred on for oesophago-gastro duodenoscopy. The referral for oesophago-gastro duodenoscopy was as a result of a mixture of initial clinical decision thinking oesophago-gastro duodenoscopy more appropriate and the food bolus being distal and the rigid scope failing. Of those initially admitted under general surgery, 12 per cent had spontaneous resolution, 86 per cent required oesophago-gastro duodenoscopy while only 2 per cent were referred to ENT for rigid oesophagoscopy. A smaller proportion were referred to the gastroenterology team, of which 11 per cent had spontaneous resolution, 88 per cent had oesophago-gastro duodenoscopy and only 1 per cent were referred on to ENT for rigid oesophagoscopy.
The mean time taken between time of admission and time of procedure was 16.2 hours for oesophago-gastro duodenoscopy and 15.6 hours in rigid oesophagoscopy.
The average distance of the bolus from the incisors was 28 cm, with the majority found at the mid or distal oesophagus (Table 3).
Site of food bolus
Complication rates were also analysed (Table 4). Of the 57 rigid oesophagoscopies, there were 8 reported complications (complication rate of 14 per cent), including 1 perforation. Of the 206 oesophago-gastro duodenoscopies performed, there were only 6 reported complications (complication rate of 3 per cent), with no known perforations.
Comparison of complications rate in oesophago-gastro duodenoscopy and rigid oesophagoscopy
Medical management
The use of Hyoscine butylbromide (also known as Buscopan) was also measured. Eight per cent of patients who were given buscopan also had a diagnosis of IHD. While 61 per cent of all patients received at least one dose of Buscopan during their admission, there was a success rate of only 18 per cent (Table 5).
Success from use of medical adjuncts
There were also patients who received magnesium sulphate, Peptac, diazepam, metoclopramide or triple-dose Buscopan; all of these patients required surgical intervention.
Discussion
The oesophagoscopy procedure for relief of a foreign body has been in existence for nearly 200 years. Its historical origin in ENT is probably owed to Chevalier Jackson, who mastered the rigid oesophagoscopy approach in the early 1900s.Reference Boyd5 Prior to this, it was seen as a potentially fatal procedure in untrained hands, and even now the procedure has significant risks. Easy access to oesophago-gastro duodenoscopy has introduced an arguably safer alternative to removal of soft food bolus. However, patient pathways have not been updated to mirror this change in practice, exemplified by the high rate of initial ENT referrals, over a third, despite ensuing subsequent onward referral for oesophago-gastro duodenoscopy.
The vast majority of patients presenting with a soft food bolus were ultimately managed with therapeutic oesophago-gastro duodenoscopy. Longstreth et al. found that oesophago-gastro duodenoscopy permitted disimpaction in 98 per cent of soft food boluses.Reference Longstreth, Longstreth and Yao6 In Scotland, most oesophago-gastro duodenoscopies are performed by both general surgery and gastroenterologists. A logical framework for dealing with soft bolus impaction is needed that is pragmatic, workable and evidence based.
Furthermore, acute ENT services in the participating Scottish health boards are centralised. Patients may require inter-hospital transfer from the presenting emergency department to the acute ENT site, which adds delay to treatment, increases inconvenience and adds to the resource burden. General surgery acute teams generally are not centralised, negating the need for inter-hospital transfer.
The complication rate of oesophago-gastro duodenoscopy was found to be only 3 per cent compared to 14 per cent in rigid oesophagoscopy. The increased complication rate of rigid oesophagoscopy compared with oesophago-gastro duodenoscopy has been shown in previous studies.Reference Gmeiner, von Rahden, Meco, Hutter, Oberascher and Stein4 In addition to this, oesophago-gastro duodenoscopy is almost exclusively performed under sedation, removing the risks and complications related to a general anaesthetic and the post-operative monitoring required.Reference Hazen, Larsen and Martin7 Patients should be offered a less risky procedure wherever possible.
It was found that 71 per cent of food boluses were located in the mid or distal oesophagus. It is widely acknowledged that complications of rigid oesophagoscopy increase the further the distance traversed, therefore full oesophagoscopy is now rarely performed.Reference Wennervaldt and Melchiors8 Furthermore, if an oesophageal perforation were to happen, this is often managed by general surgery.Reference Chirica, Kelly, Siboni, Aiolfi, Riva and Asti9
Although some ENT surgeons can perform trans-nasal oesophagoscopy, it is not part of the core ENT training and practice in the UK, nor is flexible nasendoscopy (FNE) designed for use in the oesophagus. As such, the role of FNE in these patients is often redundant, as it rarely changes management.
Eosinophilic oesophagitis is a recognised predisposing factor for food bolus obstruction, being found in 26.3–54.8 per cent of patients.Reference Desai, Stecevic, Chang, Goldstein, Badizadegan and Furuta10,Reference Ntuli, Bough and Wilson11 The diagnosis of eosinophilic oesophagitis requires endoscopic evaluation and multiple random biopsies.Reference Spergel, Dellon, Liacouras, Hirano, Molina-Infante and Bredenoord12 While eosinophilic oesophagitis can be assessed with rigid oesophagoscopy, it is not a condition routinely managed under ENT.
Patient perception of the location of the food bolus (e.g. above or below the manubrium) is often used by emergency department physicians to predict the location of the food bolus and then refer to ENT or general surgery depending on its location. However, there is little evidence to suggest patient perception has any accuracy in predicting the location of the food bolus.Reference Cook13,Reference Ko and Enns14
We were unable to collect data on patient localisation of the food bolus as documentation was often not specific enough regarding localisation. However, most food bolus were located in the mid to distal oesophagus, and the high ENT referral is likely due to inaccurate localisation to the throat and misunderstanding of the pathophysiology by referring clinicians. It could be postulated that once sedation or anaesthesia is administered, the bolus may migrate distally because of reduced oesophageal tone. However, delayed regurgitation of fluids is perhaps a better hallmark of distal location, although no evidence was found to demonstrate this.
We found the same rate of spontaneous resolution for patients receiving Buscopan as for patients swallowing a fizzy beverage. There is very little evidence to support the use of pharmacological interventions in food bolus.Reference Leopard, Fishpool and Winter3,Reference Hardman, Sharma, Smith and Nankivell15 The current popular use of Buscopan is probably a result of the erroneous reference in Scott-Brown's Otolaryngology (6th edition) to a poorly designed and uncontrolled study of 16 patients in 1991 receiving various medical adjuncts; Buscopan was not even one of the original interventions but is recommended.Reference Scott-Brown16,Reference Tibbling and Stenquist17
• Oesophageal soft food bolus obstruction is a well-recognised and not uncommon presentation to emergency departments
• Many oesophageal soft food bolus obstructions are found to resolve spontaneously, but some cases require surgical intervention
• Very mixed practice for a single common presentation was demonstrated
• The majority of patients are already managed by surgery and gastroenterology services with good outcomes and low morbidity
• Patients presenting with soft food bolus obstruction should be referred to local surgery and gastroenterology services in the first instance
Buscopan also has side effects, and the Medicines and Healthcare products Regulatory Agency put out an alert in 2017 regarding the very small risk of serious adverse reactions in patients with underlying cardiac disease.18 A significant number of our patients had ischaemic heart disease, implying the risks outweigh any benefit in the administration of Buscopan for food bolus. The use of medical adjuncts has a very poor evidence base, and even with this study, the only potential adjunct we would advocate is fizzy drinks.
The main limitation of this study arises from its retrospective character and the uncontrolled non-standard documentation and heterogenous practices.
Conclusion
We demonstrate very mixed practice for a single common presentation. We also show that the majority of patients are already managed by surgery and gastroenterology services with good outcomes and low morbidity. We therefore advocate that patients presenting with soft food bolus obstruction should be referred to local surgery and gastroenterology services in the first instance. This would reduce the need for inter-hospital transfers, minimise delays to treatment and avoid the risks associated with rigid oesophagoscopy. Further work will be required to identify risk factors in the patients necessitating transfer to ENT services (2 per cent in this study) to make the framework as robust as possible. This will form the basis of further audit work.
Competing interests
None declared
