Neurally adjusted ventilatory assist in patients with critical illness-associated polyneuromyopathy

Diaphragmatic electrical activity (EA(di)), reflecting respiratory drive, and its feedback control might be impaired in critical illness-associated polyneuromyopathy (CIPM). We aimed to evaluate whether titration and prolonged application of neurally adjusted ventilatory assist (NAVA), which delivers pressure (P (aw)) in proportion to EA(di), is feasible in CIPM patients. Peripheral and phrenic nerve electrophysiology studies were performed in 15 patients with clinically suspected CIPM and in 14 healthy volunteers. In patients, an adequate NAVA level (NAVAal) was titrated daily and was implemented for a maximum of 72 h. Changes in tidal volume (V (t)) generation per unit of EA(di) (V (t)/EA(di)) were assessed daily during standardized tests of neuro-ventilatory efficiency (NVET). In patients (median [range], 66 [44-80] years), peripheral electrophysiology studies confirmed CIPM. Phrenic nerve latency (PNL) was prolonged and diaphragm compound muscle action potential (CMAP) was reduced compared with healthy volunteers (p < 0.05 for both). NAVAal could be titrated in all but two patients. During implementation of NAVAal for 61 (37-64) h, the EA(di) amplitude was 9.0 (4.4-15.2) mu V, and the V (t) was 6.5 (3.7-14.3) ml/kg predicted body weight. V (t), respiratory rate, EA(di), PaCO2, and hemodynamic parameters remained unchanged, while PaO2/FiO(2) increased from 238 (121-337) to 282 (150-440) mmHg (p = 0.007) during NAVAal. V (t)/EA(di) changed by -10 (-46; +31)% during the first NVET and by -0.1 (-26; +77)% during the last NVET (p = 0.048). In most patients with CIPM, EA(di) and its feedback control are sufficiently preserved to titrate and implement NAVA for up to 3 days. Whether monitoring neuro-ventilatory efficiency helps inform the weaning process warrants further evaluation.