Three RCTs (n=82) were included.
One RCT described the randomisation method. None of the studies reported double-blinding. The withdrawal rates were high in the two studies reporting this (72% and 33%). The authors assumed the use of an intention-to-treat analysis in two RCTs and a per protocol analysis in the third. One RCT reported the baseline comparability of the treatment groups, the groups were incomparable in another RCT, and the third RCT did not assess this.
Inspiratory muscle strength and endurance: none of the RCTs found any significant difference between treatments for inspiratory muscle strength.
Dyspnoea: one RCT (n=30) found that IMT significantly improved dyspnoea compared with the control (reduction in modified 10-point Borg scale: 2.9 with IMT versus 1.4 with control).
Spirometry measures: the results were mixed. One RCT (n=30) found IMT significantly improved vital capacity, total lung capacity and forced expiratory volume compared with the control. The other two RCTs (n=12 and n=40) found no significant difference in lung volume between treatments.