What’s so interesting?

De Carvalho and co. show that pre-operative voice analysis can be predictive of difficult laryngoscopy.

I’d never thought about that...

The authors describe how different frequency components and acoustic qualities of the voice are, at least partly, determined by the shape and size of different anatomical areas of the vocal tract. By analysing the most intense frequencies (voice formants) within the voice spectrum they were able to correlate components with difficult laryngoscopy, namely Cormack & Lehane grade 3 or 4.

The practicalities

During pre-anaesthetic assessment, 467 elective general surgical patients were asked to pronounce each of the five vowels, corresponding to base phonemes. This was recorded on a smartphone and then later processed and analysed on a laptop computer.¹

They found...

A model using voice ‘formants’ could reliably predict difficult laryngoscopy with a ROC-AUC of 0.761 (ie. 76% probability that it correctly classifies a patient as difficult or not). When combined with the modified Mallampati this improved to 92%.

The big picture

While interesting, it’s worth remembering that using voice formants (76%) did not perform as well as modified Mallampati alone (87%), and that this performance is also surprisingly much better than those from the most recent Cochrane meta-analysis (2018) of bedside airway assessment. Over 133 studies the Cochrane review reported a summary sensitivity of only 53% and specificity of 80% for the modified Mallampati (vs 100% and 75% respectively for this study).

Although this is an interesting and novel new test, it’s just not that simple... Screening for an uncommon outcome using tests with imperfect sensitivity and specificity is already problematic, but doubly so when we are not always certain which outcome we should be screening for (laryngoscopy, intubation, ventilation, oxygenation...).

As an airway screening test, voice analysis is both different and also more of the same.

---

1. It would also be feasible for recording, analysis and reporting to occur entirely at the bedside on a smartphone. ↩

summary

• C C de Carvalho, D M da Silva, A D de Carvalho Junior, J M Santos Neto, B R Rio, C N Neto, and F A de Orange.
• Instituto de Medicina Integral Prof. Fernando Figueira (IMIP), Recife, Pernambuco, Brazil.
• Anaesthesia. 2019 Sep 1; 74 (9): 1147-1152.

AbstractWe examined the potential for voice sounds to predict a difficult airway as compared with prediction by the modified Mallampati test. A total of 453 patients scheduled for elective surgery under general anaesthesia with tracheal intubation were studied. Five phonemes were recorded and their formants analysed. Difficult laryngoscopy was defined as the Cormack-Lehane grade 3 or 4. Univariate and multivariate logistic regression were used to examine the association between some variables (mouth opening, sternomental distance, modified Mallampati and formants) and difficult laryngoscopy. Difficult laryngoscopy was reported in 29/453 (6.4%) patients. Among five regression models evaluated, the model achieving better performance to predict difficult laryngoscopy, after a variable selection criteria (stepwise, multivariate) and included a modified Mallampati classification (OR 2.920; 95%CI 1.992-4.279; p < 0.001), first formant of /i/(iF1) (OR 1.003; 95%CI 1.002-1.04; p < 0.001), and second formant of /i/(iF2) (OR 0.998; 95%CI 0.997-0.998; p < 0.001). The receiver operating curve for a regression model that included both formants and Mallampati showed an area under curve of 0.918, higher than formants alone (area under curve 0.761) and modified Mallampati alone (area under curve 0.874). Voice presented a significant association with difficult laryngoscopy during general anaesthesia showing a 76.1% probability of correctly classifying a randomly selected patient.© 2019 Association of Anaesthetists.

This article appears in the collection: What is the best method for bedside airway assessment?.

21 keywords...

hide…