Along with the 31 RCTs included in the previous version (2006), we included 34 additional RCTs in this update, resulting in a total
of 65 RCTs involving 3822 participants for inclusion in the meta-analysis.
We noted no significant demographic differences at baseline between members of the intervention group and those who received usual
care. For the pulmonary rehabilitation group, the mean forced expiratory volume at one second (FEV1) was 39.2% predicted, and for
the usual care group 36.4%; mean age was 62.4 years and 62.5 years, respectively. The gender mix in both groups was around two
males for each female. A total of 41 of the pulmonary rehabilitation programmes were hospital based (inpatient or outpatient), 23 were
community based (at community centres or in individual homes) and one study had both a hospital component and a community
component. Most programmes were of 12 weeks’ or eight weeks’ duration with an overall range of four weeks to 52 weeks.
The nature of the intervention made it impossible for investigators to blind participants or those delivering the programme. In addition,
it was unclear from most early studies whether allocation concealment was undertaken; along with the high attrition rates reported by
several studies, this impacted the overall risk of bias.