Transoral Ultrasound Versus no Transoral Ultrasound in Diagnostic Workup of Peritonsillar Abscess

Introduction Peritonsillar abscess (PTA) is a painful infection with potentially severe complications. The incidence is estimated at 37 per 100,000 annually, with a peak in teenage life and decline until old age, making it the most common deep neck infection.

Peritonsillar cellulitis (PTC) is an inflammation of the peritonsillar area due to infection but without pus formation. To distinguish PTA, PTC, and acute tonsillitis based on symptoms and clinical presentations alone can be difficult, as they resemble. This includes fever, sore throat, trismus, dysphagia, peritonsillar swelling, and deviation of the uvula away from the affected area. Distinguishing PTA, PTC, and acute tonsillitis is important, as the treatment and diagnostic approach differ. The current treatment for suspected PTA is aspiration, incision, or quinsy tonsillectomy combined with antibiotics. In contrast, the only treatment for PTC and acute tonsillitis is antibiotics.

Computed tomography (CT) scan is a well-known image modality for distinguishing PTA, PTC, and acute tonsillitis. However, this imaging modality is expensive, exposes patients to radiation, and delays treatment. Ultrasound is a radiation-free, cost-effective, and hands-on modality. Benefits that CT scans lack. The conditions for ultrasound usage in the ENT outpatient clinic are good, and ENT physicians commonly use it in Denmark. Ultrasound to diagnose a suspected PTA can be performed cervically and transorally. Both methods are considered valuable tools in PTA diagnostics. In a 2023 meta-analysis, a study describe the use of transoral ultrasound to distinguish PTA and PTC with a transoral 91% specificity and 75% sensitivity. So far, no studies investigating transoral ultrasound of PTA have included more than 46 patients; one study compared transoral ultrasound with a relevant CT scan, and only two randomized controlled trials (RCT) compared transoral ultrasound with clinical evaluation.

Smaller transoral ultrasound probes provide direct access to the peritonsillar region through the oral cavity, yielding more distinct ultrasound imaging than cervical ultrasound. It is also possible to do transorally ultrasound-guided aspiration in contrast to the cervical ultrasound.

Methods This is a prospective multicenter cluster randomized trial reported according to the SPIRIT guidelines. Cluster randomization is relevant in this setting, as this study is a low-intervention study; it does not involve any interventions beyond transoral ultrasound, which is equally effective for diagnosing a PTA. Informed consent following the claims §§ 3-5 is irrelevant, as both randomisations are a standard way to diagnose a PTA, the disease is an acute condition, and a 24-hour consideration period is impossible. This study investigates whether transoral ultrasound improves diagnostic accuracy and outcomes in patients with suspected PTA.

Randomization Patients with suspected PTA will be randomized 1:1 to groups A and B. The cluster randomization will happen weekly using an online randomization tool. All participants who have their initial outpatient clinic visit within the same week will be randomized to the same trial arm. Information about the weekly trial arm will be available at every site.

Group A: Transoral ultrasound in combination with clinical examination. Group B: Receive a solely clinical examination. This applies only to the initial outpatient clinic visits.

Study participants and recruitment Study participants are patients referred to the ENT department for suspected PTA. The suspicion is typically raised by a general practitioner or a physician at an emergency department.

Patients are included and examined at the Department of Otorhinolaryngology, Head and Neck Surgery, at Rigshospitalet, Zealand University Hospital Køge, North Zealand Hospital Hillerød and Aarhus University Hospital.

Participant study information will be provided to patients during their outpatient clinic visit. Patients will be informed that they cannot decide on the type of diagnostics (with or without US), but they can withdraw from the study and decide whether their data should be used.

Patients can withdraw their consent at any time without affecting their ongoing treatment. Even if a patient withdraws from the study, they will receive the diagnostic evaluation in the correct randomization arm. When patients withdraw their consent, their data will not be included in the study, and it will be noted in their medical chart. The patient's medical chart will be accessed once by the study coordinator to confirm that the patient does not wish to participate and that their data will not be used. This happens when the study coordinator gathers all data from patients' medical charts for data processing.

Patients will be included on the day of their outpatient clinic clinical examination. Patients are included during their first visit to the outpatient clinic. During subsequent visits, clinicians can examine patients as they prefer.

The specific data to be collected and written in the medical chart is :

• Cluster trial arm
• Is ultrasound possible? Yes/no.
• Initial diagnosis (before aspiration attempt)
• Treatment
• Transoral ultrasound experience of head and neck surgeon
• Aspirations attemps, number
• Was pus retrieved
• ultrasound characteristics

The investigators will review the medical charts of all patients with a suspected PTA seen at the participating centers during the inclusion period. This includes patients with peritonsillar cellulitis, acute tonsillitis, mononucleosis, and peritonsillar abscess. The investigators will only use data from patients who accept study participation. Among the specific data collected, the investigators will retrieve the following:

• Number of hospital visits.
• Number of admissions and timeperiod.

• Treatment, including

  • Total number of aspiration attempts and aspirations with pus, and how they are performed (with or without ultrasound)
  • Antibiotics
  • Number of surgeries (tons a chaud).
  • Objective findings: trismus, unilateral protrusion, uvula deviation, resistance upon palpation. Findings during fiberoptic examination.
  • Symptoms: unilateral predominance.
• Initial diagnosis after first visit and final diagnosis

Training of ENT residents at all departments The ENT residents will be introduced to transoral ultrasound before the study. They will participate in a didactic presentation about the execution of transoral ultrasound of patients with a suspected PTA, followed by a hands-on session. The hands-on session will focus on performing transoral ultrasound and identifying relevant anatomical landmarks. The ENT residents will also be introduced to the inclusion procedure as a part of the didactic presentation.

Statistics Power calculation The power calculation is based on the primary outcome: diagnostic accuracy of transoral ultraosund compared to solely clinical evaluation.

Data from the 12 studies that investigate transoral ultrasound of PTA patients is pooled and results in a 79.4% diagnostic accuracy.

Most studies investigating transoral ultrasoundof PTA do not compare transoral ultrasound with clinical assessment, but instead focus solely on its diagnostic accuracy. Two studies have compared transoral ultrasound with clinical assessment in PTA diagnostics and found a diagnostic accuracy of clinical evaluation of 64.2% and 64%.

A power calculation comparing two independent binomial proportions, a diagnostic accuracy of transoral ultrasound (79.4%), and a sole clinical evaluation (64.2%). The power is 85% and the significance level is 5%, yielding 136.4 patients per trial arm.

As this is a cluster-randomized study, the investigators must consider inter-cluster variation. The investigators estimate the inter-cluster variation (ICC) to be small (0.05). The investigators estimate a 5% rate og patients unable to participate in a full transoral ultrasound scan. This results in a patient cohort of 368.

Inclusion period The population of the Capital Region and Region Zealand is 2.7 million people. Aarhus University Hospital and Aalborg University Hospital cover an estimated 1.3 million people. An estimate that 37/100.000 inhabitants have a PTA annually. The investigators estimate that 25% of PTA patients are under 18 years old, 50% of PTAs are treated by private head and neck practitioners, and 50% of patients referred actually have a PTA. This results in 1110 eligible patients annually. The inclusion period will be from the first of June 2025 to the 1st of June 2026, but some centers will start including later than others.

Statistical analysis Measures of diagnostic accuracy for groups A and B will be calculated and compared, including the sensitivity, specificity, positive and negative predictive values, and overall accuracy. The presence of a PTA is defined as the aspiration of at least 0.5 mL of pus at the current visit or the following visit, no later than 4 days after the first visit. Aspiration without pus at the initial visit, but with pus one of the following two days, will be interpreted as a correct diagnosis of PTA at the initial visit. The absence of pus on aspiration or recovery with conservative treatment is defined as the patient not having a PTA.

As this is a cluster-randomized study, the investigators will use a random-effects model to assess intra-cluster variation and whether observations are independent.

Descriptive statistics for groups A and B will be generated and compared. The following differences between group A and group B will be compared using Fischer's exact test: The diagnostic accuracy, the number of patients having a tons a chaud, the number of patients receiving needle aspiration during their first visit to the outpatient clinic, and the number of visits to the outpatient clinic.

A binary logistic regression will be used to compare the number of needle aspirations during the initial visit in the outpatient clinic. This will also include patients where transoral ultrasound-guided aspiration is performed. It will also be used to evaluate if clinicians are better at diagnosing patients when they have performed more US scans.

Subgroup analysis will be performed based on symptoms and objective findings (including trismus) to determine whether transoral ultrasound would be beneficial for a particular subgroup of patients.

Discusssion Studies have shown improved diagnostic accuracy when transoral ultrasound is used to evaluate patients with suspected PTA. The investigators hypothesize that transoral ultrasound will increase the diagnostic accuracy of PTA compared to solely clinical examination, which is a visual and palpatory assessment. This will result in fewer unnecessary aspiration attempts and overlooked PTAs. The investigators aim to determine whether transoral ultrasound is a standard part of the clinical examination by comparing clinical examinations with and without ultrasound.