Dysphagia after stroke is common, especially in severely affected patients who have had a tracheotomy. In a pilot trial, pharyngeal electrical stimulation (PES) improved swallowing function in this group of patients. We aimed to replicate and extend this single-centre experience.

We did a prospective, single-blind, randomised controlled trial across nine sites (seven acute care hospitals, two rehabilitation facilities) in Germany, Austria, and Italy. Patients with recent stroke who required tracheotomy were randomly assigned to receive 3 days of either PES or sham treatment (1:1). All patients had the stimulation catheter inserted; sham treatment was applied by connecting the PES base station to a simulator box instead of the catheter. Randomisation was done via a computerised interactive system (stratified by site) in blocks of four patients per site. Patients and investigators applying PES were not masked. The primary endpoint was assessed by a separate investigator at each site who was masked to treatment assignment. The primary outcome was readiness for decannulation 24–72 h after treatment, assessed using fibreoptic endoscopic evaluation of swallowing and based on a standardised protocol, including absence of massive pooling of saliva, presence of one or more spontaneous swallows, and presence of at least minimum laryngeal sensation. We planned a sequential statistical analysis of superiority for the primary endpoint. Interim analyses were to be done after primary outcome data were available for 50 patients (futility), 70 patients, and every additional ten patients thereafter, up to 140 patients. Analysis was by intention to treat. This trial is registered with the ISRCTN registry, number ISRCTN18137204.

From May 29, 2015, to July 5, 2017, of 81 patients assessed, 69 patients from nine sites were randomly assigned to receive PES (n=35) or sham (n=34) treatment. Median onset to randomisation time was 28 days (IQR 19–41; PES 28 [20–49]; sham 28 [18–40]). The Independent Data and Safety Monitoring Board recommended that the trial was stopped early for efficacy after 70 patients had been recruited and primary endpoint data for 69 patients were available. This decision was approved by the steering committee. More patients were ready for decannulation in the PES group (17 [49%] of 35 patients) than in the sham group (three [9%] of 34 patients; odds ratio [OR] 7·00 [95% CI 2·41–19·88]; p=0·0008). Adverse events were reported in 24 (69%) patients in the PES group and 24 (71%) patients in the sham group. The number of patients with at least one serious adverse event did not differ between the groups (ten [29%] patients in the PES group vs eight [23%] patients in the sham group; OR 1·30 [0·44–3·83]; p=0·7851). Seven (20%) patients in the PES group and three (9%) patients in the sham group died during the study period (OR 2·58 [0·61–10·97]; p=0·3059). None of the deaths or serious adverse events were judged to be related to PES.

In patients with stroke and subsequent tracheotomy, PES increased the proportion of patients who were ready for decannulation in this study population, many of whom received PES within a month of their stroke. Future trials should confirm whether PES is beneficial in tracheotomised patients who receive stimulation similarly early after stroke and explore its effects in other cohorts.

Phagenesis Ltd.