Mechanical insufflation-exsufflation (MI-E) is used to augment cough in children with neurodisability. The optimal settings are unknown. We aimed to determine the superiority between three MI-E strategies regarding comfort and cough flows.

This multicentre, crossover, randomised trial was conducted at four regional hospitals in Norway. Children with neurodisability using long-term MI-E via mask were enrolled and, in a randomised order, tested three different MI-E settings strategies:

(In-/exsufflation pressure [cmH2O] / In vs. exsufflation time)

A-symmetric (±50, In=Ex)

B-asymmetric (+25-30)/-40, In>Ex)

C-personalised - as set by their therapist

The primary outcomes were user-reported comfort on a Visual Analog Scale (VAS)(0=max) and peak cough flow (PCF)(l/min) measured by a pneumotachograph in the MI-E circuit. VAS comfort scores of <4, 4-6, 7-10 was defined as mild moderate or strong discomfort, respectively. PCF of 160 l/min was used as therapeutic threshold in cildren above 12 years.

We recruited 74 children mean (SD) age 9.4 (5.3) years (range 0.6 to 17.9), and analysed 218 MI-E sequences. The mean (SD) VAS comfort score was 4.7 (2.96), 2.9 (2.44) and 3.2 (2.46) for strategies A, B and C, respectively (A versus B and C, p=0.0001). The mean (SD) PCF registered during strategies A, B and C were 203 (46.87), 171 (49.44), and 166 (46.05) l/min, respectively (A vs B and C, p<0.0001).

An asymmetric or personalised MI-E treatment strategy resulted in superior comfort scores, but lower peak cough flows than a symmetric approach utilising symmetric high pressures. All three strategies generated cough flows above the terapeutic threshold. The children rated the MI-E as slightly to moderately uncomfortable.