Does therapy with biofeedback improve swallowing in adults with dysphagia?: a systematic review and meta-analysis

ObjectiveTo describe and systematically review the current evidence on the effects of swallow therapy augmented by biofeedback in adults with dysphagia (PROSPERO 2016:CRD42016052942).Data sourcesTwo independent reviewers conducted searches which included MEDLINE, EMBASE, trial registries and grey literature up to December 2016.Study selectionRandomised controlled trials (RCTs) and non-RCTs were assessed, including for risk of bias and quality.Data extractionData were extracted by one reviewer and verified by another on biofeedback type, measures of swallow function, physiology and clinical outcome, and analysed using Cochrane Review Manager (random effects models). Results are expressed as weighted mean difference (WMD) and odds ratio (OR).Data SynthesisOf 675 articles, we included 23 studies (n=448 participants). Three main types of biofeedback were used: accelerometry, surface electromyography and tongue manometry. Exercises included saliva swallows, manoeuvres and strength exercises. Dose varied between 6-72 sessions for 20-60 minutes. Five controlled studies (stroke n=95; head and neck cancer n=33; mixed aetiology n=10) were included in meta-analyses. Compared to control, biofeedback augmented dysphagia therapy significantly enhanced hyoid displacement (three studies, WMD=0.22cm; 95% CI [0.04, 0.40], p=0.02) but there was no significant difference in functional oral intake (WMD=1.10; 95%CI [-1.69, 3.89], p=0.44) or dependency on tube feeding (OR =3.19; 95%CI [0.16, 62.72], p=0.45). Risk of bias was high and there was significant statistical heterogeneity between trials in measures of swallow function and number tube fed (I2 70-94%). Several non-validated outcome measures were used. Subgroup analyses were not possible due to a paucity of studies.ConclusionsDysphagia therapy augmented by biofeedback using surface electromyography and accelerometry enhances hyoid displacement but functional improvements in swallowing are not evident. However data are extremely limited and further larger well-designed RCTs are warranted