Statistical analysis
The meta-mean with 95% confidence interval (CI) was calculated based on the mean and SD of spirometry parameters. If a study only reported the median, range and/or inter-quartile range (IQR); mean and SD were estimated, according to Hozo et al (12). The Cochran Q statistic and inconsistency index (I2) were used to assess the heterogeneity among studies. If I2 was more than 50%, and p value was lesser than 0.05, heterogeneity was considered significant. The random effect model was used in significant heterogeneity, whereas the fixed effect model was applied for non-significant heterogeneity. To assess the stability of the results, sequential omitting of individual studies in the meta-analysis was performed using sensitivity analysis. Subgroups were analyzed based on disease severity. Probable confounders were verified using meta regression. The standardized mean difference (SMD) was calculated in studies, which measured spirometry parameters twice. Publication bias was assessed using Egger’s linear regression test. Agreement between authors in data selection and extraction was assessed using Cohen’s kappa statistic. Statistical analysis was performed using Comprehensive Meta-Analysis (CMA) computer program (Biostat, Englewood, NJ). A p value less than 0.05 was considered as statistically significant.