Acoustic parameters of snoring sound to assess the effectiveness of the Müller Manoeuvre in predicting surgical outcome

Objective To assess the effectiveness of the Müller Manoeuvre in predicting surgical outcome in non-apnoeic snorers. Methods Forty-one non-apnoeic snorers performed the Müller Manoeuvre, prior to palatal surgery for snoring. Pre-operatively and between 1.0 and 4.1 months (mean 2.5 months) post-operatively, patients were admitted overnight when their sleeping position and snoring sounds were recorded. At the time of the post-operative recordings, patients were required to complete a specifically designed questionnaire. Snore files comprising the inspiratory component of the first 100 snores whilst the patient was supine, were extracted. Snore duration (s), snore loudness (dBA), snore periodicity (%) and the energy ratios for the frequency bands 0–200, 0–250 and 0–400 Hz were calculated. Only patients who showed improvements in snore periodicity and all energy ratios were considered to be surgical successes. In addition, patients were also categorised as ‘successes’ or ‘failures’ depending on their responses to specific questionnaire questions. The effectiveness of the Müller Manoeuvre in predicting surgical outcome was then tested using these categories. Results The 41 patients included 35 men and 6 women. Mean age: 47 years (24–67 years). Mean PNIFR 145 (80–230). Median reported alcohol intake was 11–15 units/week (0 to 26–30 units/week). Mean BMI: 30.6 kg/m2 (24.3–47.2 kg/m2). Twenty-four patients underwent an uvulopalatal elevation palatoplasty and seventeen a traditional palatoplasty. Following the Müller Manoeuvre, patients were categorised as ‘ideal’, ‘suboptimal, but acceptable’ or ‘unsuitable’ for surgery. Using the acoustic parameters, 23/41 patients were considered a surgical success, whilst 18/41 were considered failures. Using the questionnaire responses, 14/40 patients were considered a surgical success, whilst 26/40 were considered failures. There was no correlation between the subjective and objective outcomes (ρ = 0.193; p = 0.227). Neither pre-operative BMI, type of palatoplasty performed, patient gender, age, PNIFR or reported alcohol intake were confounders of surgical outcome. For patients considered ‘ideal’ and ‘suboptimal, but acceptable’, using acoustic outcomes, the Müller Manoeuvre had a specificity of 55.5% and a sensitivity of 30.4%, compared with a sensitivity of 57.7% and a specificity of 28.6% when questionnaire responses were used. If only patients considered ‘ideal’ were considered, the specificity was 66.7%, and the sensitivity 21.7% when using acoustic outcomes, compared with a sensitivity of 69.2% and a specificity of 78.6% when questionnaire responses were used. Conclusion The Müller Manoeuvre appears to have no role in the pre-operative assessment of palatal surgery for non-apnoeic snorers