Eighteen studies were included for review (n= 815): three parallel double blind RCT (n=100), one parallel single-blind trial (n=38), five parallel unblinded trials (n=504), seven double-blind crossover trials (n=139), one randomised continuous sequential double-blind study (n=33) and one study design that was not reported. Validity was not assessed, but the authors noted that allocation generation was not reported in nine studies and allocation concealment was not reported in 11 studies.
Morphine (seven studies, n=256)
Subcutaneous morphine resulted in significantly decreased dyspnea intensity compared to placebo (two studies, n=19; no p-values reported). In one study (n=101; no p-values reported) morphine combined with midazolam improved dyspnea significantly (92 per cent dyspnea relief) compared to morphine alone (69 per cent dyspnea relief) or midazolam alone (46 per cent dyspnea relief). Nebulized morphine and morphine dosage were not associated with any significant changes in dyspnea. Three trials reported adverse events of nausea, sedation and somnolence in 33 patients.
Oxygen (six studies, n=149)
Three studies found no significant difference between oxygen and air on dyspnea intensity (n=122). However, studies that included only hypoxic patients showed significant benefits of oxygen on dyspnea intensity (two studies n=15; no p-values reported). One study reported a significant reduction in dyspnea VAS scores with helium enriched air (n=12, no p-values reported).
All three studies of nursing led breathlessness rehabilitation reported significant benefits compared to no intervention control groups (no p-values reported). Furosemide and acupuncture and acupressure did not show any significant benefit on dyspnea.