Fourteen RCTs (1,359 patients) were included in the review. Methodological quality was rated as low with a mean Jadad score of 2. One RCT scored 5, four RCTs scored 3, two RCTs scored 2, and seven RCTs scored 1. Three RCTs reported methods of randomisation; five RCTs reported double blinding, with only one reporting an appropriate method of blinding; five RCTs reported drop-outs and withdrawals, two RCTs were placebo controlled.
Herbal medicines versus placebo or no treatment (six RCTs): All studies used conventional therapy in both the intervention and control groups.
Shen Mai: Significant improvements were reported for Shen Mai injection (three herbs with Panax ginseng as the main ingredient) compared to no treatment for forced expiratory volume in one second (FEV1) (weighted mean difference 0.58, 95% CI: 0.19 to 0.97; one RCT, 38 patients), vital capacity (weighted mean difference 1.64; 95% CI: 0.98, 2.30), responder rate (risk ratio 3.90, 95% CI: 1.32 to 11.51) and the Borg scale symptom score (weighted mean difference -1.20, 95% CI: -2.02 to -0.30).
Panax ginseng: There was insufficient data to calculate an effect size for one RCT (n=92 patients) comparing Panaxginseng extract with placebo.
Salvia miltiorrhiza: A significant difference was reported for Salvia miltiorrhiza (Danshen) injection compared with no treatment for FEV1 (weighted mean difference 0.53, 95% CI: 0.36 to 0.70; one RCT, 53 patients). However no significant differences were reported between groups for responder rate.
Jiawei Yupingfeng: A significantly higher responder rate was reported for Jiawei Yupingfeng compared with no treatment (risk ratio 1.44, 95% CI: 0.80 to 2.58; one RCT, 84 patients).
Jinshui Liujin: No statistically significant differences between Jinshui Liujin decoction and no treatment were found between groups for responder rate, FEV1 or vital capacity (one RCT, 82 patients).
Echinacea: No statistically significant differences were reported for FEV1 comparing Echinacea based liquid extract (Esberitox N) to placebo (one RCT, 53 patients, text and tables differ).
Herbal medicine versus conventional therapy (one RCT):
Hedera helix versus ambroxol: No statistically significant differences were found comparing Hedera helix leaf extract (Prospan) with conventional therapy of ambroxol tablets (one RCT, n=94 patients).
Herbal medicine versus herbal medicine (seven RCTs):
Ke Chuan Ping versus Qing Jin Hua Tan: Statistically significant improvements were reported for Ke Chuan Ping decoction compared to Qing Jin Hua Tan decoction for responder rate (risk ratio 1.62, 95% CI: 1.00 to 2.61) but not for FEV1 (one RCT, 62 patients).
13 herb anti-cough dyspnoea versus Ephedra-almond: Statistically significant improvements were reported for 13 herb anti-cough dyspnoea decoction compared to Ephedra-almond decoction for responder rate (risk ratio 2.77, 95% CI: 1.96 to 3.93; one RCT, 400 patients).
Yiqi Mianyi versus Zhenqi Fuzheng: Statistically significant improvements were reported for Yiqi Mianyi granules compared to Zhenqi Fuzheng granules for responder rate (risk ratio 1.96, 95% CI: 1.15 to 3.34) and for symptom score (weighted mean difference -2.70, 95% CI: -5.57 to -0.17; one RCT, 102 patients).
Bufei Keli versus Yupingfeng Keli: Statistically significant improvements were reported for Bufei Keli granules compared to Yupingfeng Keli granules for responder rate (risk ratio 1.71; 95% CI: 1.11 to 2.64), symptom severity score (weighted mean difference -1.03, 95% CI: -2.03 to -0.03) and days of exacerbation (weighted mean difference -7.20, 95% CI: -11.50 to -2.90; one RCT, 62 patients).
Kensuning versus Jinbei Tankeqing: One RCT (n=120) comparing Kensuning granule and placebo with Jinbei Tankeqing granule found no significant differences between groups for responder rates.
Gubenhuatanquyu versus Buyiguben: Symptom scores were statistically significantly improved in a group using Gubenhuatanquyu decoction compared to a group using Buyiguben decoction (weighted mean difference -2.00, 95% CI: -3.88 to -0.12; one RCT, 65 patients).
Two forms of Hedera helix extract (Prospan): No statistically significant differences were found for Prospan cough syrup compared with Prospan herbal drops for FEV1 or vital capacity (one RCT, 50 patients).
Adverse events were reported in the review.