Introduction
An aspirated foreign body in a child can represent a potentially life-threatening emergency. The diagnosis can be challenging due to vague symptoms, unwitnessed events, subtle examination findings, minimal radiological signs, and symptoms that overlap with lower respiratory-tract infection. Many aspirated foreign bodies, such as food and plastic, are radiolucent, making them undetectable by conventional radiographic methods. Definitive diagnosis is often made by bronchoscopy under general anaesthesia.
Several clinical prediction models exist that incorporate history, clinical exam findings, and radiological signs. However, a meta-analysis by Lee et al. suggested that existing models have not been adequately validated, and none can currently be recommended to aid clinical decision-making.Reference Lee, Philteos, Levin, Namavarian, Propst and Wolter1
The typical average effective dose for a paediatric chest radiograph for a five-year-old is 0.02 mSv, equivalent to three days of exposure to natural radiation.2 However, the effective dose depends on many factors, including weight, age, and patient sex. In contrast, the typical effective dose from a standard paediatric CT chest is 3 mSv.2 Children may have a higher sensitivity to radiation-induced cancers than adults.Reference Kutanzi, Lumen, Koturbash and Miousse3 The sensitivity, specificity, and negative and positive predictive values of chest CT for assessing foreign bodies range from 91–100 per cent compared with 23–71 per cent with radiographs.Reference Kwok, Wong, Paddle, Goergen and Rimmer4, Reference Ahmed, Guillerman and Giannoni5 However, despite the superiority of CT as a diagnostic tool, it has not been integrated into routine practice due to radiation exposure concerns and the potential need for sedation.
Some recent studies have described an ultra-low dose CT introduced for selective patients to assess for an aspirated foreign body.Reference Koşucu, Ahmetoğlu, Koramaz, Orhan, Ozdemir and Dinç6, Reference Gordon, Nowik, Mobini Kesheh, Lidegran and Diaz7 An ultra-low dose protocol was introduced at this institution for this indication by fixing the kilovoltage to 70kV and Smart mA range from 10–35mA. Organ dose modulation was also utilised to reduce the dose anteriorly to the thyroid. A noise index of 30 was prescribed to allow for a noisy low-dose acquisition. Images were taken with a single axial block (100–160 mm depending on patient size) and a fast rotation time of 0.28 sec. The target field of view was limited to include the trachea and left and right main bronchi only. The scanner in our institution has a large detector, 16 cm, which gives the ability to scan a large field of view, up to 16 cm, in one axial rotation of 0.28 sec. This extremely short scan time helps reduce motion artefacts from the heart and lungs. For young children, parents wear a lead coat and remain in the scan room with them to help comfort and keep the child still.
This study compared the estimated effective radiation dose children received during ultra-low dose CT scans with that received with traditional cumulative radiographic investigations. Additionally, it analysed the differences in cost and length of hospital stay between patients who had CT and those who underwent radiographs.
Methods
This study was carried out in a tertiary referral paediatric hospital in Ireland. It received ethical approval from the local ethics and research board. This is a retrospective study from 2014 to 2024. Hospital in-patient enquiry data were collected over 10 years on all patients with a suspected airway-aspirated foreign body at this institution. Data collection included basic demographics, length of stay, radiological investigations performed, treatment received, bronchoscopy, and whether a foreign body was found at the time of bronchoscopy. Depending on the radiological investigation received, patients were divided into two groups: radiographs or ultra-low-dose CT.
At this institution, before the introduction of ultra-low dose CT, children with suspected aspirated foreign bodies are assessed in the emergency department and undergo a chest radiograph. If there was a clinical concern for an aspirated foreign body, the child would be referred for assessment by an otolaryngologist. The patient may proceed for bronchoscopy in theatre or be admitted for observation. Those admitted for observation may have repeated radiographs including expiration/decubitus views at times and/or undergo a bronchoscopy during their admission if there are ongoing clinical concerns. Since introducing ultra-low dose CT, stable patients with suspected aspirated foreign bodies have had a CT performed in the emergency department. If the CT is negative for a foreign body, the patients are discharged, or if the CT is positive, the patients go directly to theatre for bronchoscopy.
Radiographs were performed using X-ray equipment programmed with dedicated paediatric protocols. Systems used included Siemens FD-X (Munich), Siemens Multitom Rax (Munich), Carestream DRX-Evolution Plus (Rochester, NY), and a Philips DigitalDiagnost (Hamburg). The low-dose CT scans were performed on a GE Revolution CT scanner (Chicago) using a purpose-built protocol for suspected foreign body ingestion, outlined in the introduction.
The effective doses were estimated using two different software packages. PCXMC 2.0 (Helsinki) was used for the radiographic exams, while NCICT 3.0 (Bethesda, MD) was used to calculate the CT effective dose. The cumulative effective dose from the multiple radiographs that patients received during their hospital stay, including neck, chest, and abdominal radiographs, was compared with the single estimated effective dose of an ultra-low dose CT.
Consultant paediatric otolaryngologists performed bronchoscopy in theatre under general anaesthesia. The cost was calculated using costs provided by the institutional finance department. The cost of admission to the hospital was €304, the cost of a CT scan was €304, a radiograph was €120, the cost per night in the hospital was €850, and the cost of bronchoscopy in theatre was €1100.
Statistical analysis was carried out using Jamovi (version 2.0.0.0) for Macintosh (Sydney, Australia). The Mann–Whitney U-Test was used to calculate statistical cost differences, length of stay, and effective doses between groups.
Results
Forty-four patients were included in the study: 32 in the radiograph group and 12 in the CT group. The median age in the CT group was 43.5 months (3.6 years), ages 17–144 months, while in the radiograph group the median age was 17 months (1.4 years), ages 4–180 months. The median length of stay in the CT group was 0 days, range 0–3 days. The median length of stay in the radiograph group was 1 day, range 0–17 days. Within the radiograph group, four patients required re-admission to hospital. All of these patients underwent a bronchoscopy in theatre, and all had a foreign body at the time of bronchoscopy. There were no re-admissions in the CT group.
The median cost in the CT group was €304, costs in this group ranged from €304 to €1258, compared with a median cost of €2419 in the radiograph group, ranging from €240 to €16,334. In the radiograph group the median cumulative estimated effective dose was 0.0178 mSv, with values of 0.003–0.086 mSv. The median estimated effective dose in the CT group was 0.075 mSv, estimated effective doses in this group were 0.04–0.23 mSv. Required data to estimate the effective dose were missing for 17 radiographs, which could not be included in the analysis (Table 1).
Descriptive data of included patients; LOS = length of stay
There was a statistically significant difference in the length of stay and cost in the CT group when compared with the radiograph group (p < 0.01), with the CT group having shorter length of stay and lower cost. There was a statistically significant difference in the cumulative estimated effective radiation dose in the radiograph compared to the estimated effective dose received in the CT group (p < 0.01), with the radiograph group having a lower estimated effective dose. No patients required sedation for CT imaging (Table 2; Figures 1–5).
Flowchart of patients in radiograph group; ED = emergency department; LOS = length of stay.
Flowchart of patients in the CT group.
Plot of difference in mean cost in €.
Plot of difference in mean dose in mSv.
Plot of difference in mean length of stay (LOS) in days.
Statistical analysis using Mann–Whitney U test between CT and radiograph groups for dose, length of stay (LOS), and cost
Twenty-nine foreign bodies were found at the time of bronchoscopy. Two patients, both in the radiograph group, underwent bronchoscopy where no foreign body was found. Three patients had foreign bodies found by CT. The report matched the location of the foreign body at the time of bronchoscopy in all three patients. Foreign bodies found at the time of bronchoscopy included peanut (n = 12), apple (n = 3), bead (n = 3), sunflower seed (n = 3), LEGO® (n = 2), plastic (n = 2), pistachio (n = 1), carrot (n = 1), wood (n = 1), and tooth (n = 1). The right main bronchus was the most common site of the foreign body (n = 16), followed by the left main bronchus (n = 6), and next most common sites were the bronchus intermedius (n = 4) and trachea (n = 2) (Table 3; Figures 6 and 7).
CT Image demonstrating no evidence of a foreign body in the airway (coronal view).
CT image demonstrating a peanut in the bronchus intermedius (coronal view); foreign body annotated with *.
Analysis of foreign body found at the time of bronchoscopy
Discussion
Prompt diagnosis of suspected aspirated foreign bodies in the airway is necessary to prevent morbidity and mortality. Clinical and radiological findings are often non-diagnostic, and bronchoscopy is required for confirmation. The negative bronchoscopy rate for suspected foreign bodies in the airway is 20–50 per cent.Reference Cavel, Bergeron, Garel, Arcand and Froehlich8, Reference Kadmon, Stern, Bron-Harlev, Nahum, Battat and Schonfeld9 Multiple clinical risk-prediction models have been proposed to assist in diagnosis. However, a recent meta-analysis found a high risk of bias and inadequate validation among proposed models. Thus, no current models can be recommended to guide decision-making.Reference Lee, Philteos, Levin, Namavarian, Propst and Wolter1
Chest radiographs are often performed as the first investigation and are frequently normal. Most foreign bodies are radiolucent, thus indirect signs are what is present on radiographs. These include air trapping, atelectasis, and rarely pneumothorax. The radiographic signs can vary over time; some are only observed later. During the first 24 hours, the radiograph has a low sensitivity (62–88 per cent) and specificity (30–97 per cent) and is reported to be normal in 14–37 per cent of cases.Reference Hitter, Hullo, Durand and Righini10 However, its benefits include being widely available, cheap, low radiation exposure, and not requiring sedation.
CT as a diagnostic tool is superior to radiographs and has reported detection rates for a bronchial foreign body of close to 100 per cent.Reference El Khoury, Makhoul, El Hadi, Haber and Rassi11 This modality also provides valuable information to the surgeon, such as the location and size of the foreign body. Gordon et al. compared effective doses used in ultra-low dose CT with conventional radiographic methods (fluoroscopy and chest radiography) and found that doses in the CT group were lower.Reference Gordon, Nowik, Mobini Kesheh, Lidegran and Diaz7 The median dose for CT was 0.04 mSv compared with 0.1 mSv for conventional methods. The CT group also had higher sensitivity, specificity, and positive and negative predictive values.Reference Gordon, Nowik, Mobini Kesheh, Lidegran and Diaz7
Concerns regarding ionising radiation in the paediatric population exist. Despite the superiority of CT to radiograph as a diagnostic tool, concerns for the radiation dose received by the child during CT have meant CT is not being used as a first-line investigation.
Another factor to consider is the potential need for sedation or general anaesthetic in children undergoing CT. In this study no patients required sedation or anaesthesia for CT. There is a small risk of long-term cognitive deficits in children who undergo anaesthesia, in particular before the age of three years.Reference Ing, DiMaggio, Whitehouse, Hegarty, Brady and von Ungern-Sternberg12 Of greatest concern, however, is the risk of adverse respiratory events in children with upper respiratory infections who undergo general anaesthesia. The incidence of adverse respiratory events in this cohort is as high as 30 per cent.Reference Michel, Vacher, Julien-Marsollier, Dadure, Aubineau and Lejus13 As discussed previously, children with foreign body aspiration or upper respiratory infection may present with a similar clinical picture. In the radiograph group in this study, two patients underwent a bronchoscopy that showed no foreign body. With the use of CT, diagnostic bronchoscopy under anaesthesia in children with suspected foreign body aspiration can be avoided if the CT is negative. Avoiding anaesthesia is particularly useful in the cohort with an upper respiratory infection mimicking foreign body aspiration.
This study found a statistically significant reduction in cost and length of stay when CT was used to assess for a foreign body in the airway rather than a radiograph. However, the limitations of this study include a small sample size in the CT group. Hospital in-patient enquiry data coding is imperfect, and some cases may not have been identified due to incorrect diagnosis coding in the 10 years of this study (i.e. patients in whom foreign body aspiration was in the initial differential diagnosis, but ultimately were diagnosed with a respiratory tract infection). A large number of bronchoscopies were performed in the radiograph group—87.5 per cent compared with 25 per cent in the CT group. This suggests that patients in the radiograph group may have been positively selected from hospital in-patient enquiry coding (i.e. there are an unknown number of patients who may have initially been treated as a suspected foreign body aspiration but their ultimate discharge code in hospital in-patient enquiry does not reflect this and thus they were not captured in our data).
Computed tomography can be used as first-line investigation in the emergency department for stable patients with suspected foreign body aspiration. This can facilitate earlier discharge of patients with a negative CT who are clinically well and identify those who need foreign body removal in theatre. The ultra-low dose CT protocol can be replicated by other units, which may reduce the number of patients transferred to a tertiary centre with a suspected foreign body aspiration diagnosed on history, examination, and radiograph.
• Ultra-low dose computed tomography (CT) is a cost-effective and safe diagnostic tool in the investigation of paediatric foreign body aspiration
• No children required sedation or general anaesthetic for CT
• Chest radiographs are often performed as the first investigation and are frequently normal
• Ultra-low dose CT should be performed as the first and only investigation
Conclusion
Ultra-low dose CT is a safe, cost-effective first-line investigation in stable patients with suspected foreign body aspiration.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
The authors declare none.
