Oral administration of P. communis has a significant antipyretic effect in mice with fever induced by dried yeast. The mechanism of this antipyretic effect is associated with the inhibition of IL-1β, TNF-α, and cAMP expression in vivo, as well as the suppression of cyclooxygenase in the hypothalamic temperature center, ultimately reducing the release of PGE2 (Liu et al., 2021).

Stigmasta-3,5-dien-7-one, derived from P. communis, exhibits potent anti-inflammatory activity by suppressing the lipopolysaccharide-stimulated production of NO, PGE2, and cytokines (TNF-α, IL-1β, and IL-6), as well as inhibiting the induction of iNOS and COX-2 protein in lipopolysaccharide-induced RAW 264.7 cells (Park et al., 2016).

The hydrolysis of P. communis-derived oligosaccharides (ROs) using H₂O₂ has shown significant antibacterial activity, with a 13.57 mm inhibition zone against S. aureus at a concentration of 100 μg/mL (Qian and Jiang, 2014).

Additionally, active ingredients extracted from the above-ground parts of P. communis using methanol, petroleum ether, and carbon tetrachloride have demonstrated strong peripheral analgesic activity in Swiss albino mice (Sultan et al., 2017).

Furthermore, P. communis has been found to enhance immunity. Gavage administration of fresh P. communis aqueous extract increased lymphocyte transformation, NK cell activity, and T-cell immune response function in mice, with more pronounced effects at higher doses (22.5 g/kg) (Sun et al., 2016; Zhang et al., 2016). However, this is a very high dose and is unlikely to be achievable in humans.

Although these findings seem promising, they are all from laboratory experiments (in vitro and in animals). It is very important to understand the effectiveness of P. communis preparations in human patients because this is key to deciding whether they can be recommended in clinical treatment guidelines. There has been a notable absence of systematic reviews examining the clinical evidence about the efficacy, effectiveness, and safety of P. communis in the treatment of ARTIs. This review aims to fill this gap to inform the development of evidence-based clinical guidelines for the treatment of common infections.

We searched eight electronic databases up to 8 February 2023: China National Knowledge Infrastructure (CNKI), Chinese Scientific Journal Database (VIP), Chinese BioMedical Literature Database (CBM), Wanfang Database (Wanfang), PubMed, the Cochrane Library, Embase, and Web of Science. Additionally, we conducted searches on ClinicalTrials.gov and the Chinese Clinical Trial Registry from their inception to 8 February 2023.

Our search terms were adapted to match the requirements of each database. These search terms encompassed “Phragmites communis,” “Lu Gen,” and “acute respiratory tract infection,” among others. Further details on additional search terms and strategies in both Chinese and English, tailored to each specific database, can be found in Supplementary Table S1. To identify additional relevant studies, we also reviewed the references of eligible articles. We imposed no language restrictions, and translations were procured whenever necessary.

Four authors (MF, LYK, FLP, and GHJ) independently screened the titles and abstracts of all potential studies identified through our searches. When there was uncertainty, insufficient information, or cases of disagreement, we obtained the full texts of articles and then determined eligibility from the full texts. Reasons for excluding articles at the full-text screening stage were recorded. After identifying eligible studies, two authors independently carried out data extraction using a planned data extraction form. In case of a disagreement, it was resolved through negotiation involving another author (JPL). For the included trials, we extracted the following data, as recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Lefebvre et al., 2022):

1) General information: title, first author, publication language, publication year, country, and settings; 2) participants: diagnosis, symptom duration, total number enrolled, and the number in each comparison group, along with baseline characteristics; 3) interventions: herbal CONSORT items, including the herbal medicinal product name, characteristics of the herbal product, quality control, dosage regimen, and quantitative description; 4) follow-up: length of follow-up, reason for dropouts and withdrawals, and the number of participants affected; 5) outcomes reported: mean and standard deviation (SD) for continuous outcomes and the number of events for dichotomous outcomes; and 6) adverse events.

MF, LYK, YFL, and YZS were responsible for the risk of bias assessment. For each trial, two authors independently assessed the risk of bias using the Cochrane Collaboration Risk of Bias tool 2.0 (Higgins et al., 2022). Disagreements were discussed and resolved regarding the original protocol and, if necessary, arbitration by another author (XYH).

The Risk of Bias Tool 2.0 is a structured tool for assessing the risk of bias in RCTs. It is designed to guide systematic reviewers in conducting meaningful, outcome-based assessments of trial design, implementation, and reporting. The tool is divided into five fixed bias domains: bias arising from the randomization process, bias due to deviations from intended interventions, bias due to missing outcome data, bias in the measurement of the outcome, and bias in the selection of the reported result. By using this tool, reviewers can identify potential sources of bias and evaluate the overall reliability of the study.

We used RevMan 5.3 software for data analysis. Risk ratios (RRs) with corresponding 95% confidence intervals (CIs) were computed for dichotomous variables, while mean differences (MDs) with 95% CI were calculated for continuous data. In cases where outcome measures had consistent units across studies, we reported the effects as standardized mean differences (SMDs).

RevMan 5.3 software is a piece of software dedicated for creating and managing Cochrane systematic reviews. It provides a user-friendly interface for conducting meta-analyses, generating forest plots, and presenting the results in a clear and concise manner.

In the trials, we separated the arms into different comparisons to avoid double-counting participants with multiple intervention groups that met the inclusion criteria.

We proactively reached out to investigators or authors to validate critical study details and obtain missing numerical outcome data when needed (e.g., when a study reported outcomes with a line chart). Where standard deviation was not reported by means, it was calculated from the information reported, such as CIs or p-values. When we did not get a response, we only used the available data in the analyses.

We assessed between-study heterogeneity using the I ² statistic, which quantifies the percentage of variation across studies attributed to heterogeneity rather than chance. The rule of thumb for the interpretation of this statistic indicated that I ² > 30% signifies moderate heterogeneity, I ² > 50% indicates substantial heterogeneity, and I ² > 75% suggests considerable heterogeneity. When I ² values were above 50% for primary outcomes, we conducted a sensitivity analysis to explore potential sources of heterogeneity and factored these findings into our interpretation of the results.

We conducted Egger tests using R software version 4.2.2 to explore potential reporting bias, where applicable, and when a sufficient number of studies were available within a single meta-analysis.

Where possible, we planned to conduct our analyses based on intention-to-treat (ITT) data for each outcome, which include data from all randomized participants in the individual trials, regardless of any deviations from the original study plan. When possible, we extracted the end-of-treatment scores for continuous outcomes rather than relying on the change-from-baseline score. Given the anticipated variability in the populations and interventions across the included trials, we used a generic inverse variance random-effects model to pool the data, allowing for the incorporation of heterogeneity into our analysis.

Subgroup analyses were performed for the primary outcomes where there were sufficient studies in each comparison group:

1. P. communis as a monotherapy versus being part of an herbal mixture

We performed sensitivity analyses for the primary outcome to determine whether the conclusions would change when limiting eligibility to trials with a low risk of overall bias. In cases of significant heterogeneity, we planned to perform sensitivity analysis to explore potential sources of heterogeneity in more detail.

A total of 8,820 records were obtained. After removing 794 duplicates, 8,026 records remained and were screened by title and abstract. Then, 463 records were assessed in full text. A total of 421 studies were excluded for not meeting the inclusion criteria as they were not RCTs (n = 8), did not involve adults with ARTIs (n = 159), did not contain Lu Gen in the intervention group or contained TCM in the control group (n = 253), or had wrong data (n = 1). Finally, 42 trials involving 6,879 participants, published between 2006 and 2022, met the inclusion criteria and were included for meta-analysis (Zhang, 2006; Wang, 2007; Guo, 2009; Huai, 2009; Shi et al., 2010; Wang, 2010; Hui et al., 2012; Wong et al., 2012; Xue, 2012; Xu and Zhi, 2013; Zhou, 2013; Sun, 2014a; Sun, 2014b; Chen, 2014; Pan, 2014; Wei, 2014; Xu, 2014; Chang, 2016; Chen and Li, 2016; Cheng et al., 2016; Zhang, 2016; Zhang et al., 2016; Guo and Xie, 2017; Xue et al., 2017; Zhang, 2017; Bi et al., 2019; Hu, 2019; Luo and Zhong, 2019; Yang, 2019; Zheng, 2019; Hong et al., 2020; Li, 2020; Wang, 2020; Ye et al., 2020; Song, 2021; Zhou and Qu, 2021; Li et al., 2022a; Li et al., 2022b; Li et al., 2022c; Li et al., 2022d; Sun, 2022). Two trials were published in English, and the remaining were in Chinese (Figure 1). Figure 1 shows the PRISMA flowchart and trial inclusion.

The included studies were conducted in patients with AECOPD (n = 13) (Song, 2021; Li, 2020; Zhang, 2017; Wei, 2014; Li et al., 2022a; Li et al., 2022b; Zheng, 2019; Zhang et al., 2016; Chen, 2014; Shi et al., 2010; Hong et al., 2020; Wang, 2020; Chang, 2016), acute bronchitis (n = 2) (Bi et al., 2019; Ye et al., 2020), influenza (n = 6) (Wang, 2007; Hui et al., 2012; Sun, 2014b; Zong and Chen, 2020; Li et al., 2022c; Li et al., 2022d), acute upper respiratory infections (n = 21) (Zhang, 2006; Guo, 2009; Huai, 2009; Wang, 2010; Wong et al., 2012; Xue, 2012; Xu and Zhi, 2013; Zhou, 2013; Sun, 2014a; Pan, 2014; Xu, 2014; Chen and Li, 2016; Cheng et al., 2016; Zhang, 2016; Guo and Xie, 2017; Xue et al., 2017; Hu, 2019; Luo and Zhong, 2019; Yang, 2019; Zhou and Qu, 2021; Li et al., 2022c), and acute respiratory infection with fever (Sun, 2022). A total of 16 studies (Zhang, 2006; Huai, 2009; Wong et al., 2012; Xu and Zhi, 2013; Zhou, 2013; Sun, 2014a; Sun, 2014b; Pan, 2014; Cheng et al., 2016; Xue et al., 2017; Yang, 2019; Zong and Chen, 2020; Li et al., 2022a; Li et al., 2022c; Li et al., 2022d; Sun, 2022) included patients with onset within 48 h. A total of 13 studies (Wang, 2007; Guo, 2009; Shi et al., 2010; Wang, 2010; Xue, 2012; Chen, 2014; Wei, 2014; Xu, 2014; Chen and Li, 2016; Bi et al., 2019; Hu, 2019; Luo and Zhong, 2019; Zhou and Qu, 2021) included patients with the onset of illness within 3–23 days, and 13 trials (Hui et al., 2012; Chang, 2016; Zhang, 2016; Zhang et al., 2016; Guo and Xie, 2017; Zhang, 2017; Zheng, 2019; Hong et al., 2020; Li, 2020; Wang, 2020; Ye et al., 2020; Song, 2021; Li et al., 2022b) did not report the details of onset.

The mixtures containing P. communis included Chinese patent medicines (n = 7) (Huai, 2009; Wong et al., 2012; Xue, 2012; Zhou, 2013; Pan, 2014; Sun, 2014a; Li et al., 2022c), fixed Chinese herbal medicine formulas (n = 25) (Wang, 2007; Shi et al., 2010; Xu and Zhi, 2013; Sun, 2014a; Sun, 2014b; Chen, 2014; Wei, 2014; Xu, 2014; Chang, 2016; Cheng et al., 2016; Zhang, 2016; Zhang et al., 2016; Xue et al., 2017; Zhang, 2017; Bi et al., 2019; Hu, 2019; Yang, 2019; Zheng, 2019; Li, 2020; Wang, 2020; Zong and Chen, 2020; Song, 2021; Li et al., 2022d; Li et al., 2022b; Sun, 2022), and individualized Chinese herbal medicine formulas (n = 11) (Zhang, 2006; Guo, 2009; Wang, 2010; Hui et al., 2012; Chen and Li, 2016; Guo and Xie, 2017; Luo and Zhong, 2019; Hong et al., 2020; Ye et al., 2020; Zhou and Qu, 2021; Li et al., 2022a).

Six trials (Wong et al., 2012; Zhou, 2013; Cheng et al., 2016; Xue et al., 2017; Yang, 2019; Li et al., 2022c) compared P. communis versus the placebo. Twenty trials (Shi et al., 2010; Wang, 2010; Sun, 2014b; Chen, 2014; Wei, 2014; Xu, 2014; Zong and Chen, 2020; Chang, 2016; Zhang et al., 2016; Zhang, 2017; Bi et al., 2019; Zheng, 2019; Hong et al., 2020; Li, 2020; Wang, 2020; Ye et al., 2020; Song, 2021; Li et al., 2022a; Li et al., 2022b; Sun, 2022) compared P. communis plus usual care versus usual care. Sixteen trials (Zhang, 2006; Wang, 2007; Guo, 2009; Huai, 2009; Hui et al., 2012; Xue, 2012; Xu and Zhi, 2013; Sun, 2014a; Pan, 2014; Chen and Li, 2016; Zhang, 2016; Guo and Xie, 2017; Hu, 2019; Luo and Zhong, 2019; Zhou and Qu, 2021; Li et al., 2022d) compared P. communis versus usual care. Seven (Guo, 2009; Hui et al., 2012; Xue, 2012; Sun, 2014a; Xu, 2014; Li et al., 2022c; Li et al., 2022b) of the included studies allowed adjuvant treatments, such as oral acetaminophen, if the patient had a temperature ≥39°C.

Three included studies (Hui et al., 2012; Xue, 2012; Song, 2021) reported the time to disappearance of the main symptoms (cure time). A total of 36 studies (Zhang, 2006; Wang, 2007; Guo, 2009; Shi et al., 2010; Wang, 2010; Hui et al., 2012; Xu and Zhi, 2013; Zhou, 2013; Sun, 2014a; Sun, 2014b; Chen, 2014; Pan, 2014; Wei, 2014; Xu, 2014; Chang, 2016; Chen and Li, 2016; Cheng et al., 2016; Zhang, 2016; Zhang et al., 2016; Guo and Xie, 2017; Xue et al., 2017; Zhang, 2017; Bi et al., 2019; Hu, 2019; Luo and Zhong, 2019; Yang, 2019; Zheng, 2019; Hong et al., 2020; Li, 2020; Ye et al., 2020; Zong and Chen, 2020; Song, 2021; Zhou and Qu, 2021; Li et al., 2022a; Li et al., 2022d; Sun, 2022) reported the proportion of patients with symptoms resolved (cure rate). Seven studies (Sun, 2014a; Zhang, 2006; Xu, 2014; Cheng et al., 2016; Zhang, 2016; Yang, 2019; Zong and Chen, 2020; Sun, 2022) reported on the cooling onset time. Ten studies (Zhang, 2006; Hui et al., 2012; Xue, 2012; Sun, 2014a; Xu, 2014; Cheng et al., 2016; Zhang, 2016; Yang, 2019; Zong and Chen, 2020; Sun, 2022) reported on the fever clearance time. However, only few studies have reported definitions of the cooling onset time and the fever clearance time. Three studies (Zhang, 2006; Cheng et al., 2016; Sun, 2022) defined the cooling onset time as the time it takes for the body temperature to decrease by 0.5°C after the initiation of medication. Two studies (Hui et al., 2012; Sun, 2014b) defined the fever clearance time as the time it takes for the medication to be administered until the body temperature drops to normal and can be maintained for more than 24 h, whereas three studies (Zhang, 2006; Cheng et al., 2016; Sun, 2022) defined the fever clearance time as the time from the administration of the drug until the body temperature drops to normal. The rest of the studies did not give detailed information.

No trials reported time spent absent from school or work due to illness. Twelve studies (Zhou, 2013; Sun, 2014a; Sun, 2014b; Wei, 2014; Zhang et al., 2016; Guo and Xie, 2017; Zhang, 2017; Li, 2020; Li et al., 2022a; Li et al., 2022b; Li et al., 2022d; Sun, 2022) reported TCM syndrome scores about the CMCRG. The primary and secondary symptoms of the scales differed slightly in each study due to the specific disease. The reported scales all included respiratory symptoms, mainly cough, sputum, runny nose, nasal congestion, fever, sore throat, and physical discomfort. There was also some variation in the scores for each symptom across the scales.

Sixteen studies (Zhang, 2006; Huai, 2009; Wong et al., 2012; Zhou, 2013; Sun, 2014b; Pan, 2014; Wei, 2014; Cheng et al., 2016; Zhang et al., 2016; Zhang, 2017; Wang, 2020; Ye et al., 2020; Zhou and Qu, 2021; Li et al., 2022a; Li et al., 2022d; Li et al., 2022c) reported adverse events. Some studies also focused on lung function and inflammation-related indicators. All the outcomes were measured during or after treatment. The duration of treatment was 2 days to 4 weeks.

Seven (Zhou, 2013; Sun, 2014a; Sun, 2014b; Wei, 2014; Zhang, 2017; Li, 2020; Sun, 2022) of 42 reports were degree theses. More information about the characteristics of the included studies is provided in Supplementary Table S2.

Seven studies (Sun, 2014a; Wong et al., 2012; Zhou, 2013; Cheng et al., 2016; Xue et al., 2017; Yang, 2019; Li et al., 2022a) used Chinese patent medicines, including four listed medicines, namely, Ganmao Qingre granules, antiviral oral liquid, Siji antiviral mixture, and Yin Qiao San, and two in-hospital preparations: Qingre Huashi oral liquid and Qingjie Kanggan granules. Ganmao Qingre granules, an antiviral oral liquid, and Yin Qiao San are included in the Pharmacopoeia of the People’s Republic of China 2020. The pharmacopoeia provides a detailed record of all the ingredients and the exact manufacturing process for these medicines. Detailed information is provided in Supplementary Table S3.

Twenty-six trials used fixed Chinese herbal medicine formulas and reported their composition. Only five of these trials reported the origin of the TCM compound prescriptions, three of which were from classical compound prescriptions, i.e., The Essentials of the Golden Chamber, The Essentials of Thousand Gold, and The Woman’s Good Prescription, and the other two were from the empirical prescriptions of Prof. J.J. Chen and Prof. Q.N. Wan, respectively. No studies reported on the authentication of each ingredient. Twenty-five studies reported on the rationale for the composition of the herbal compound. Only eight studies reported pharmacological or toxicological related tests. Detailed information is provided in Supplementary Table S4.

Eleven trials used individual Chinese herbal medicine formulas and reported the composition and principle of the formulation. Only one trial showed Phragmitis rhizoma as an assistant medicinal. One study did not report the method of preparation. The rest of the formulations were either decocted in water or supplied directly by the pharmacy. No studies were reporting on the safety assessment of Chinese medicine compounding. Detailed information is provided in Supplementary Table S5.

The authentication methods, chemical analysis, and quality control of each ingredient were not reported for Chinese patent medicines, fixed Chinese herbal medicine formulas, and individual Chinese herbal medicine formulas.

We selected one outcome from each trial that we considered important to evaluate. One of the 42 studies was at low risk of bias; 3 studies had some concerns; and 38 studies were assessed as high risk. Detailed information is shown in Figure 2.

A total of 42 studies involving 6,879 participants were included in the systematic review of P. communis, all conducted in China, with 40 published in Chinese and 2 in English. None of the included studies used P. communis as monotherapy. Thus, the specific role played by P. communis in ARTIs is not clear.

The route of administration for all P. communis combinations is oral. P. communis preparations improved the cure rate and shortened the time to fever reduction in each of the three control conditions. The combination of P. communis preparation with usual care significantly improved lung function and inflammatory response (except for CRP and NEUT%) in patients with AECOPD or acute bronchitis. The P. communis preparation significantly increased the clearance rate of influenza virus nucleic acid. However, the number of participants included was too small, and the results may be unstable.

Sixteen studies reported adverse events, and the results showed no statistically significant difference in the incidence of adverse reactions between the P. communis group and the control group.

The methodological quality of the included studies was poor, mainly in terms of non-reporting of the randomization method and allocation concealment in most studies, non-reporting of blinding, lack of sufficient information to judge deviation from the intended intervention, and bias in the measurement of the outcome. Study protocols were found in only two of the included articles, so it is also important to register study protocols before the study and to have a pre-specified analysis plan before starting the analysis. The number of authors for 23 studies was one, and the lack of acknowledgment of other researchers in the text casts doubt on the credibility of the original studies and suggests that multidisciplinary collaborative clinical trials should be conducted in the future. The quality of evidence for all outcomes was assessed using GRADE and was found to be of low or very low certainty.

The studies included have hardly reported the authentication, chemical analysis, and quality control of Chinese herbal medicine, which is detrimental to the reproducibility of experiments. This is because we know that the composition of Chinese herbal medicine may vary depending on the origin, time, and batch of the medicine. Therefore, it is difficult for us to judge the consistency of experimental validity with the herbal compound used.

This study is the first systematic review to date evaluating P. communis for the treatment of ARTIs. We conducted an extensive search in multiple databases and with no language restriction to be as comprehensive as possible.

Studies selected were restricted to adults aged 18 years and older and restricted dosing to oral administration, so we cannot comment on effects in children or alternative preparations. No studies were identified that used P. communis as a single herb, and the role played by P. communis in the various formulas is not clear, so we are unable to determine if the observed findings are specific to P. communis. In addition, Chinese medicine compound formulas consist of a wide range of herbs, among which herbs that are often paired with P. communis may have effects similar to those of P. communis, which makes the effects of P. communis even more difficult to isolate. The included studies were conducted in China, so the conclusions drawn are regionally limited.

As there are not enough studies of identical Chinese medicine compound prescriptions, we combined different compound prescriptions together for analysis. Admittedly, the effects of these herbal compound prescriptions are not identical. In addition, we analyzed different types of ARTIs, such as influenza and cold, acute bronchitis, and AECOPD, which have similar symptoms but are not treated in exactly the same way. It is possible that P. communis is more effective for some types of ARTIs than for others.

P. communis is distributed in several regions, including China, Japan, Korea, Europe, and North America. However, most studies have focused mainly on the role of P. communis in the environment, while its medicinal potential remains largely neglected. To our knowledge, there is no published systematic review of P. communis for ARTIs. P. communis is often used as an assistant medicinal in herbal compound formulas in TCM. Only one formula with P. communis as a monarch medicine was included in this review. The results showed that the formula including P. communis could alleviate the overall symptoms of ARTIs and shorten the time to fever reduction, which is consistent with the previously published findings of pharmacology experiments (Qian and Jiang, 2014; Zhao et al., 2022) and animal experiments with P. communis (Liu et al., 2021; Park et al., 2016; Sun et al., 2016, Liu and Liang, 2014).

Although P. communis is often used as an assistant medicinal in TCM formulas, this review included a variety of TCM formulas containing P. communis for ARTIs, which suggests that P. communis plays an important role and is well worth further in-depth study.

A total of 118 herbal medicines were included in this review. The herbs that were most frequently combined with P. communis were Platycodonis Radix (29 times), Glycyrrhizae Radix et Rhizoma (24 times), Armeniacae Semen Amarum (19 times), Forsythiae Fructus (18 times), Menthae Haplocalycis Herba (16 times), and Scutellariae Radix (15 times). Future studies may further explore the use of P. communis in combination with these herbs to understand how they work together.

The results showed that P. communis preparations had significant efficacy in improving symptoms in patients with ARTIs, especially in febrile patients and patients with AECOPD. The results suggest that P. communis preparations improve the clearance rate of influenza virus nucleic acid, but due to the small number of participants included, a larger study is recommended to further validate this result in the future. It is also important to identify which preparations are most effective for which types of ARTIs. On the basis of current evidence, this is not possible because there were many preparations and few trials of each. There were no studies that used antibiotic use as an outcome, so future studies could explore whether the use of P. communis preparations could reduce antibiotic use.

There are currently four Chinese patent medicines containing P. communis: Ganmao Qingre granules, antiviral oral liquid, Siji antiviral mixture, Yin Qiao San, and an in-hospital preparation of Qingjie Kanggan granules that can be used directly in the clinical setting. This review found that P. communis preparations were effective in alleviating the symptoms of ARTIs and could improve the cure rate. Therefore, P. communis preparations have the potential to be used as an alternative symptomatic treatment for ARTIs.