만성기침

Objectives: Chronic cough is a frequent condition worldwide that significantly impairs quality of life. Herbal medicine (HM) has been used to treat chronic cough due to the limited effectiveness of conventional medications. This study aimed to summarize and determine the effects of HM on patients with chronic cough.

Methods: A comprehensive search of 11 databases was conducted to find randomized controlled clinical trials (RCTs) that reported the effects of HM for patients with chronic cough on 16 March 2023. The primary outcome was cough severity, and the secondary outcomes included cough-related quality of life, cough frequency, total effective rate (TER), and cough recurrence rate. The methodological quality of the included studies was assessed using the Cochrane risk of bias tool, and the certainty of the evidence for effect estimates was assessed using the Grading of Recommendations, Assessment, Development, and Evaluations tool.

Results: A total of 80 RCTs comprising 7,573 patients were included. When HM was used as an alternative therapy to conventional medication, there were inconsistent results in improving cough severity. However, HM significantly improved cough-related quality of life and TER and significantly lowered the cough recurrence rate compared with conventional medication. When used as an add-on therapy to conventional medication, HM significantly improved cough severity, cough-related quality of life, and TER and significantly lowered the recurrence rate. In addition, HM had a significantly lower incidence of adverse events when used as an add-on or alternative therapy to conventional medication. The subgroup analysis according to age and cause of cough also showed a statistically consistent correlation with the overall results. The certainty of the evidence for the effect of HM was generally moderate to low due to the risk of bias in the included studies.

Conclusion: HM may improve cough severity and cough-related quality of life, and lower the cough recurrence rate and incidence of adverse events in patients with chronic cough. However, due to the high risk of bias and clinical heterogeneity of the included studies, further high-quality placebo-controlled clinical trials should be conducted using a validated and objective assessment tool.

Systematic Review Registration: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42023418736, CRD42023418736.

3.1 Study selection and characteristics: A total of 2,958 articles were identified through the database search. After removing 546 duplicate articles, 1813 articles were excluded through the title and abstract review. As a result, 597 studies underwent full-text review, except for one study where the full text was not retrieved. After the full-text review, 517 studies were excluded for the following reasons: not an RCT (n = 444), not about chronic cough (n = 30), not reporting cough period (n = 33), not about HM (n = 1), using EATM as the control group intervention (n = 6), not reporting the outcome of interest (n = 2), and duplicate data (n = 1) (Supplementary Material S2). Finally, a total of 80 studies involving 7,573 participants were included in this review (;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ) (Figure 1).

One study was conducted in Korea and published in English; the others were conducted in China and published in Chinese. The study setting of one study was a family planning service station, and the remaining settings were clinics. A total of 21 studies reported funding sources (;;;;;;;;;;;;;;;;;;;; ). A total of 27 studies were conducted on children (;;;;;;;;;;;;;;;;;;;;;;;;;; ), 3 studies included both children and adults (;; ), and the remaining 50 studies were conducted on adults. Only 30 studies (37.5%) described a cause for the cough, including post-respiratory infection in 5 studies (;;;; ); GERD (;; ), UACS (;; ), and unexplained chronic cough (;; ) in 3 studies each; and chronic bronchitis, CVA, and nonspecific chronic chough in 1 study each. There were also 13 studies targeting a mixed population, including patients with various causes of the cough (;;;;;;;;;;;; ). One study was a 3-arm RCT comparing HM plus conventional medication, HM, and conventional medication. The remaining studies were 2-arm RCTs, 33 comparing HM and conventional medication (;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; ), 1 comparing HM and placebo HM, and the rest comparing HM plus conventional medication and conventional medication alone (Supplementary Material S3).

In the included studies, various HMs were used, and among them, Zhisou-san or its modified form was used the most frequently (13 studies) (;;;;;;;;;;;; ). A total of 154 botanical drugs were used, and as a result of analyzing the constituent botanical drugs of basic HM, four studies were excluded that used different prescriptions according to the pattern identification (;;; ). Among them, Glycyrrhiza glabra L. [Fabaceae; Glycyrrhizae Radix et Rhizoma] was used the most (61 studies), followed by Platycodon grandiflorus (Jacq.) A.DC. [Campanulaceae; Platycodonis Radix] (41 studies), Aster tataricus L.f. [Asteraceae; Asteris Radix et Rhizoma] (36 studies), Pinellia ternata (Thunb.) Makino [Araceae; Pinelliae Tuber] (29 studies), Citrus × aurantium f. deliciosa (Ten.) M. Hiroe [Rutaceae; Citri Unshius Pericarpium] (28 studies), Prunus armeniaca L. [Rosaceae; Armeniacae Semen] (27 studies), Poria cocos Wolf [Polyporaceae; Poria Sclerotium] (24 studies), Stemona tuberosa Lour. [Stemonaceae; Stemonae Radix] (22 studies), and Schisandra chinensis (Turcz.) Baill. [Schisandraceae; Schisandrae Fructus] (20 studies). As for the dosage of HMs, decoction was the most common (68 studies), followed by granule (5 studies) and oral liquid (3 studies). The HM administration duration varied from 1 week to 3 months, with 2 weeks being the most (34 studies), followed by 4 weeks (13 studies), 1 month (9 studies), and 1 week (7 studies). Twelve studies performed follow-ups after the end of the HM administration, and the follow-up period varied from 1 week to 1 year (Supplementary Material S4).

3.2 Risk of bias assessment: All the studies appropriately generated random sequences using random number tables or dice-throwing methods, and none of the studies mentioned allocation concealment. Except for one study comparing HM and placebo HM, no study conducted blinding of the participants and personnel. In addition, only two studies blinded the outcome assessors (; ). Four studies were evaluated as having a high risk of attrition bias by a per-protocol analysis (;;; ). All the studies were rated as having a low risk of reporting bias and other biases (Supplementary Material S5).

3.3 HM versus conventional medication: Although there was no significant difference in the SCS total score between the two groups (MD -0.07, 95% CI −0.31 to 0.18, I2 = 0%), each daytime (MD −0.34, 95% CI −0.59 to −0.08, I2 = 91%) or nighttime SCS score (MD −0.38, 95% CI -0.58 to −0.19, I2 = 86%) significantly decreased in the HM group compared with the conventional medication group (Table 1).

Cough-related quality of life, assessed by the LCQ total score, and physical, psychological, and social sub-items, were significantly increased after the treatment in the HM group compared with the conventional medication group (total: MD 2.32, 95% CI 0.04 to 4.61, I2 = 96%; physical items: MD 0.63, 95% CI 0.30 to 0.95, I2 = 75%; psychological items: MD 0.58, 95% CI 0.30 to 0.87, I2 = 71%; social items: MD 0.66, 95% CI 0.14 to 1.18, I2 = 88%). In addition, the TER, calculated according to the cough symptom improvement, was significantly higher in the HM group compared with the conventional medication group (RR 1.23, 95% CI 1.18 to 1.29, I2 = 35%), regardless of subgroup according to age. The funnel plot for the TER was asymmetric, and the result of the Egger’s test also suggested the risk of publication bias with a p-value of less than 0.05 (Supplementary Material S6). Cough recurrence rate was significantly lower in the HM group (RR 0.29, 95% CI 0.16 to 0.51, I2 = 0%) (Table 1).

Eighteen studies comparing HM and conventional medication reported the incidence of adverse events (;;;;;;;;;;;;;;;;; ). The incidence of adverse events was significantly lower in the HM group compared with the conventional medication group (RR 0.26, 95% CI 0.17 to 0.40, I2 = 0%). The funnel plot was symmetric, and the p-value of the Egger’s test was 0.917, suggesting that the risk of publication was low (Supplementary Material S6). When analyzing subgroups according to age, insignificant results were shown between the two groups in children, but adverse events were significantly lower in the HM group in adults and mixed populations (Table 1).

3.4 HM plus conventional medication versus conventional medication alone: Cough severity, evaluated by SCS daytime and nighttime scores, was significantly lower in the HM plus conventional medication group (SCS daytime: MD −0.53, 95% CI −0.67 to −0.39, I2 = 97%; SCS nighttime: MD −0.55, 95% CI −0.69 to −0.41, I2 = 95%), which was consistent with subgroup analyses according to age and cause of cough. In addition, cough severity, assessed by the 0–100 mm cough VAS, also significantly improved in the HM plus conventional medication group (MD −11.69, 95% CI −21.71 to −1.67, I2 = 98%) (Table 1).

Cough-related quality of life, assessed by the LCQ, was significantly improved in the HM plus conventional medication group compared with conventional medication alone (physical items: MD 0.88, 95% CI 0.71 to 1.05, I2 = 0%; psychological items: MD 0.62, 95% CI 0.46 to 0.77, I2 = 0%; social items: MD 0.81, 95% CI 0.44 to 1.17, I2 = 71%), and these results were consistent in the subgroup analysis according to age. One study targeted adults with chronic cough due to GERD. When HM was additionally administered with conventional medication (rabeprazole and mosapride citrate) for 8 weeks, cough severity and quality of life, assessed by the chronic cough impact scale, were significantly improved compared to conventional medication alone (p < 0.01). One study targeted pediatric patients with chronic cough due to UACS. When HM was additionally administered with conventional medication (budesonide nasal spray) for 12 weeks, cough frequency per day was significantly improved compared to conventional medication alone (p < 0.05).

The TER, according to the cough symptom improvement, was significantly higher in the HM plus conventional medication group (RR 1.18, 95% CI 1.16 to 1.21, I2 = 0%). The funnel plot for the TER was asymmetric, and the result of Egger’s test also suggested the risk of publication bias with a p-value of less than 0.05 (Supplementary Material S6). The cough symptom recurrent rate was significantly lower in the HM plus conventional medication group (RR 0.30, 95% CI 0.19 to 0.46, I2 = 0%), and the statistical significance was consistent when conducting a subgroup analysis according to age (Table 1).

Seventeen studies comparing HM plus conventional medication and conventional medication alone reported the incidence of adverse events (;;;;;;;;;;;;;;;; ), and it was found that adverse events were significantly lower in the HM plus conventional medication group (RR 0.46, 95% CI 0.25 to 0.87, I2 = 50%) (Table 1). The funnel plot was symmetric, and the p-value of Egger’s test was 0.433, suggesting the risk of publication was low (Supplementary Material S6).

3.5 HM versus placebo HM: Only one study compared HM versus placebo HM, and therefore, a meta-analysis was not possible. In the study, HM significantly improved cough severity, measured by the cough diary score and cough VAS, and cough-related quality of life, measured by LCQ, compared with the baseline. HM significantly improved daytime and total cough diary score compared to the placebo HM.

3.6 Certainty assessment: The certainty of the evidence for all meta-analysis outcomes comparing HM to conventional medication was moderate to low. The main reason for downgrading was the risk of bias of the included studies, inconsistency, imprecision, and publication bias. The certainty of the evidence for cough severity and cough-related quality of life when comparing HM plus conventional medication and conventional medication alone was moderate, downgraded by the risk of bias of the included studies. In this comparison group, the certainty of the evidence for TER and cough recurrence rate was low, and the certainty of the evidence for adverse event rates was very low (Table 1).