Limits of effective cough-augmentation techniques in patients with neuromuscular disease.

BACKGROUND Manual and mechanical cough-augmentation techniques can improve peak cough flow (PCF) in patients with respiratory insufficiency caused by neuromuscular disease. METHODS We studied cough-augmentation techniques in 179 clinically stable patients with various neuromuscular diseases. We measured vital capacity (VC), maximum expiratory pressure (MEP), and PCF, with and without 3 cough-augmentation techniques: manually assisted cough (MAC); breath-stacking (in a subgroup of 60 patients receiving noninvasive mechanical ventilation); and breath-stacking in combination with MAC (also in the 60-patient subgroup). We analyzed the data with the receiver operating characteristic (ROC), to predict the lower limits (assisted PCF > or = 180 L/min) and upper limits (assisted PCF < unassisted PCF) of effectiveness of the 3 cough-augmentation techniques. RESULTS The lower limit of effective assisted cough with MAC, breath-stacking, and breath-stacking plus MAC was best predicted by VC > 1,030 mL (ROC 0.86, P < .001), VC > 558 mL (ROC 0.92, P < .001), and VC > 340 mL (ROC 0.90, P < .001). The upper limit of effective MAC was best predicted by MEP > 34 cm H(2)O (ROC 0.89, P < .001), whereas the ROC prediction of the upper limit of effective cough with breath-stacking and with breath-stacking plus MAC was not better than random. With each of the cough-augmentation techniques the benefits decreased linearly with increasing MEP and VC (P < .001). Compared to MAC and breath-stacking alone, breath-stacking plus MAC best improved unassisted PCF (P < .001). CONCLUSIONS In clinically stable patients with neuromuscular diseases, the effectiveness of cough-augmentation techniques can be predicted with measurements of maximum respiratory capacity. Patients with VC > 340 mL and MEP < 34 cm H(2)O would optimally benefit from the combination of breath-stacking plus manually assisted cough to improve PCF to > 180 L/min.