Dysphagia Knowledge Hub — 吞嚥困難知識庫

Cervical Auscultation as a Dysphagia Screening Tool

Cervical auscultation (CA) refers to the use of a stethoscope or electronic microphone placed on the lateral neck — typically over the thyroid cartilage or cricoid cartilage — to capture sounds produced during swallowing. Its appeal as a bedside screening tool lies in its non-invasiveness, low cost, and immediate availability. However, its evidence base is more nuanced than its widespread clinical use might suggest, and its role relative to instrumental assessments — FEES and VFSS — requires careful framing. This article reviews the acoustic characteristics of swallowing, the published evidence on diagnostic accuracy, the combination with pulse oximetry, training requirements, and the appropriate place of CA in HK clinical practice.

Acoustic Characteristics of Normal Swallowing

A normal swallow produces a characteristic acoustic signature that experienced clinicians learn to recognise. Three components have been identified in the literature (Hamlet et al., 1990; Zenner et al., 1995):

  1. Opening click: a brief transient sound at the initiation of the pharyngeal swallow, thought to correspond to rapid laryngeal elevation and cricopharyngeal opening
  2. Swallowing sound (bolus passage): a sustained, somewhat turbulent sound of variable duration as the bolus passes through the pharynx and upper oesophagus; in healthy adults this lasts approximately 300–600 ms for liquid boluses
  3. Closing click: a second transient sound corresponding to laryngeal descent and cricopharyngeal closure

Abnormal swallowing sounds associated with aspiration or penetration include: wet, gurgling, or bubbly quality during or after the swallow; prolonged bolus passage sounds; multiple swallows per bolus; coughing; and absence of the normal click-swallow-click sequence.

Clinical Application

In practice, SLTs place a standard Littmann-style stethoscope on the anterolateral neck, just lateral to the thyroid cartilage. The patient is asked to swallow standardised bolus volumes (typically 3 ml, 5 ml, and 10 ml of water, and a teaspoon of yogurt or pudding for comparison across consistencies). The clinician listens for quality changes, timing abnormalities, and post-swallow voice quality (wet voice on phonation of /a/ after the swallow — the “wet voice” sign).

The technique requires quiet surroundings, correct stethoscope placement, and training in sound interpretation. Electronic accelerometers have been explored as research tools to generate quantifiable acoustic waveforms and reduce inter-listener variability, but have not yet entered routine HK clinical practice.

Sensitivity and Specificity: Meta-Analytic Evidence

The most comprehensive systematic review and meta-analysis of cervical auscultation’s diagnostic accuracy is Lagarde et al. (2016), published in Dysphagia. The meta-analysis pooled data from 19 studies (n = 744 participants) using VFSS or FEES as the reference standard.

Key findings (Lagarde et al., 2016):

The moderate sensitivity and specificity values indicate that cervical auscultation alone is insufficient to confirm or exclude aspiration with confidence. False-negative rates of approximately 27% mean that a meaningful proportion of aspirating patients would be missed if CA were used as the sole screening instrument. False-positive rates of 28% carry the risk of unnecessary dietary restriction.

An earlier systematic review by Leslie et al. (2004) reached similar conclusions, noting that inter-rater reliability is only moderate (kappa 0.40–0.60 in most studies) and that training substantially affects accuracy. Softness of aspirated material and the presence of silent aspiration — where no cough or acoustic change is produced — are major limitations.

Combination with Pulse Oximetry

Several studies have examined whether combining CA with pulse oximetry (SpO2 monitoring) improves diagnostic accuracy over either technique alone.

The rationale is that aspiration of material into the lower airways may produce a measurable drop in SpO2, providing a physiological correlate of the acoustic finding. The criterion most commonly used is a 2% or greater desaturation from baseline during or within two minutes of swallowing.

Collins and Bakheit (1997) found that combined CA plus oximetry had higher sensitivity than oximetry alone in stroke patients, though specificity remained modest. However, Colodny (2000) and other authors found poor specificity of SpO2 desaturation for aspiration, as transient desaturation occurs for many non-aspiration-related reasons in acutely ill patients (positional change, secretion mobilisation, anxiety).

The current consensus is that combined CA plus oximetry can modestly improve sensitivity at the cost of reduced specificity, and may be appropriate as a first-line bedside screen when FEES is not immediately available — provided that positive screens are followed by instrumental assessment before dietary decisions are made.

Training Requirements

Cervical auscultation is a learned perceptual skill. Studies by Stroud et al. (2002) and Borr et al. (2007) have shown that untrained listeners cannot reliably distinguish aspirating from non-aspirating swallows, and that even trained SLTs show only moderate inter-rater reliability without systematic training programmes.

Recommended training components include:

No standardised CA certification exists in Hong Kong. The Hospital Authority Speech Therapy Service includes CA among core clinical competencies, but formal competency assessment varies by cluster.

Limitations vs. FEES and VFSS

CA cannot directly visualise the bolus or the larynx, cannot measure penetration-aspiration scale scores, cannot differentiate between penetration and aspiration, and cannot assess pharyngeal residue or cricopharyngeal function. It is unable to detect silent aspiration unless postswallow voice quality is assessed — and even “wet voice” has a sensitivity of only approximately 40–70% for aspiration (Warms & Richards, 2000).

FEES provides direct visualisation of secretion management, bolus flow, and post-swallow residue. VFSS allows real-time dynamic assessment including bolus timing, laryngeal elevation, and oesophageal phase. Both are significantly superior to CA for clinical decision-making in patients with confirmed or suspected dysphagia.

HK Clinical Practice

In Hong Kong’s public hospital system, cervical auscultation is routinely incorporated into the Clinical Swallowing Examination (CSE) conducted by SLTs, typically as part of the bedside assessment prior to FEES or VFSS referral. It is not used as a substitute for instrumental assessment in complex cases. Ward-based nursing dysphagia screens (e.g., the Toronto Bedside Swallowing Screening Test, TOR-BSST, or similar tools used at Queen Mary Hospital) do not typically include CA, as reliability by non-SLT staff has not been demonstrated.

CA is most defensible in HK as a component of the initial clinical screen that stratifies patients into low-risk (may trial oral feeding with monitoring) and higher-risk (requires instrumental assessment before oral intake) categories — with all higher-risk patients receiving FEES or VFSS.

References