Table 2. Meta-analysis of BODE Index at end of treatment and sensitivity analysis.
| CHM plus RP groups versus the RP groups | BODE score (MD+95%CI) Model |
| Pooled data for two studies of CHM+Sal./Flu.[37], [60] | −0.70 [−1.13, −0.27], I2 = 0% FEM* |
| −0.70 [−1.13, −0.27], I2 = 0% REM* | |
| Pooled data for four 3-mth studies[37], [49]–[51] | −0.76 [−1.28, −0.24], I2 = 66% FEM* |
| −0.83 [−1.81, −0.16], I2 = 66% REM | |
| Pooled data for two 6-mth studies[40], [47] | −1.12 [−1.69, −0.54], I2 = 0%FEM* |
| −1.12 [−1.69, −0.54], I2 = 0%REM* | |
| Pooled data for 5 studies of AM formulae[37], [40], [49], [58], [60] | −0.54 [−0.89, −0.19], I2 = 55.5% FEM* |
| −0.51 [−1.07, −0.05], I2 = 55.5% REM | |
| Pooled data for all 9 studies[37], [40], [42], [47], [49], [50], [54], [58], [60] | −0.71 [−0.94, −0.47], I2 = 40% FEM* |
| −0.72 [−1.06, −0.39], I2 = 40% REM* | |
| Pooled data for 8 studies[37], [40], [42], [47], [49], [50], [58], [60] after excluding the English-language study[54] | −0.68 [−0.91, −0.44], I2 = 27% FEM* |
| −0.67 [−0.97, −0.38], I2 = 27% REM* | |
| Pooled data for 8 studies exclusively indexed in Chinese databases [37], [40], [42], [47], [49], [50], [58], [60] | −0.68 [−0.91, −0.44], I2 = 27% FEM* |
| −0.67 [−0.97, −0.38], I2 = 27% REM* | |
| Pooled data for 7 studies[37], [40], [42], [47], [49], [50], [60] after excluding studies of pulmonary rehabilitation[54], [58] | −0.72 [−0.97, −0.48], I2 = 20% FEM* |
| −0.73 [−1.02, −0.44], I2 = 20% REM* | |
| Pooled data for 6 studies[37], [40], [42], [47], [58], [60] after excluding studies of undefined RP[49], [50], [54] | −0.72 [−0.97, −0.47], I2 = 2% FEM* |
| −0.72 [−0.98, −0.46], I2 = 2% REM* | |
| Pooled data for 3 studies in which mean BODE was greater than 4 at baseline[37], [40], [54] | −1.40 [−2.12, −0.69], I2 = 0% FEM* |
| −1.40 [−2.12, −0.69], I2 = 0% REM* | |
| Pooled data for 5 studies in which mean BODE was less than 4 at baseline[47], [49], [50], [58], [60] | −0.59 [−0.90, −0.29], I2 = 47% FEM* |
| −0.53 [−0.98, −0.07], I2 = 47% REM* |
*Significant difference, CHM: Chinese Herbal Medicine, RP: routine pharmacotherapy, MD: mean difference, CI: confidence interval, mth: month, REM: random effect model, FEM: fixed effect model, AM: Astragalus membranaceus, Salm./Flu.: salmeterol/fluticasone (inhaled).