Atractylodes macrocephala-Paeonia lactiflora Class Formula for the Treatment of Irritable Bowel Syndrome: A Systematic Review With Meta-Analysis and Trial Sequential Analysis

Previous meta-analyses suggested that Chinese herbal medicine (CHM) is effective for irritable bowel syndrome (IBS). Formulas with Atractylodes macrocephala and Paeonia lactiflora as the core pairs have been widely used by traditional Chinese medicine (TCM) practitioners for the treatment of IBS. We aimed to examine the efficacy and safety of the Atractylodes macrocephala-Paeonia lactiflora class formula (A-P CHM) for IBS through a meta-analysis and trial-sequential analysis (TSA). The protocol is registered in the International Prospective Register of Systematic Reviews (PROSPERO) under registration number CRD42023439087. We searched seven databases for data up to May 23, 2023. The primary outcome was global IBS symptom relief. The secondary outcomes included the IBS severity scoring system (IBS-SSS) score and treatment-related adverse events. The relative ratio (RR) (dichotomous variables), the standardized mean difference (SMD) (continuous variables), the number needed to treat (NNT), the number needed to harm (NNH), and the required information size (RIS) were calculated. Twenty-four eligible articles with 3,768 participants were included. Thirteen trials were at low risk of bias (RoB). Compared with placebo or Western medication, A-P CHM was associated with a significantly higher proportion of relief of global IBS symptoms. The TSA analysis verified the primary outcome. For the secondary outcome, the A-P CHM IBS-SSS score was lower than Western medication or placebo at the end of the treatment, which was further confirmed by the TSA analysis. We asserted that A-P CHM might be a potential candidate for patients with IBS, especially for IBS-D. It may provide a theoretical basis for future optimization of irritable bowel syndrome with diarrhea (IBS-D) herbal formulas. The overall certainty of the evidence was not high; more tightly designed randomized controlled trials (RCTs) are required in the future.


Introduction And Background
Irritable bowel syndrome (IBS) is a chronic functional bowel disorder characterized by abdominal pain linked to disordered defecation [1].Affecting 5%-10% of the global population, IBS is notably more prevalent in women and younger individuals [2].Although the pathogenesis of IBS remains poorly understood, the prevailing scholarly consensus posits abnormal regulation of the brain-gut axis (BGA) resulting from a complex interplay of factors.These include increased visceral sensitivity, gastrointestinal motility disorders, low-grade intestinal inflammation, gut microflora imbalance, psychosomatic stress, and immune activation [3].Many treatments are proposed for IBS [4][5][6], and dietary interventions and pharmacological treatments are usually the primary choices [5,7].Due to the complex pathophysiology of IBS, dietary interventions and pharmacological treatments are only partially effective, yielding marginal improvements in comparison to placebo [7,8].In instances where conventional pharmacological treatments fail to alleviate IBS symptoms, patients often resort to complementary and alternative medicines (CAMs) such as herbal medicine and acupuncture [9].
Formulas featuring Atractylodes macrocephala and Paeonia lactiflora as key components are widely employed by traditional Chinese medicine (TCM) practitioners for addressing disorders such as diarrhea, intestinal tinnitus, and spleen deficiency.Analysis of the TCM application data from the Chinese Formulas Database revealed a total of 1,557 formulas incorporating Atractylodes macrocephala and Paeonia lactiflora pairs [15,16].Modern clinics similarly utilize compound formulas with Atractylodes macrocephala and Paeonia lactiflora as the core pair for treating irritable bowel syndrome, demonstrating substantial clinical efficacy [17].Atractylodes macrocephala and Paeonia lactiflora were first published in Shennong Ben Cao Jing (Divine Husbandman's Classic of the Materia Medica).Atractylodes macrocephala is sweet, bitter, warm, and bitterwarm to dry dampness and strengthen the spleen and stop diarrhea; Paeonia lactiflora is bitter, sour, slightly cold, and sour-cold to drain fire, suppress wood, and pacify the liver [18].A pertinent study indicated that the Atractylodes Macrocephala-Paeonia lactiflora (A-P CHM) pair surpassed Atractylodes Macrocephala or Paeonia lactiflora alone in the treatment of IBS by more effectively reducing the levels of 5hydroxytryptamine, vasoactive intestinal peptide, and growth inhibitors, stabilizing mast cells, diminishing intestinal hypersensitivity, and enhancing fecal water content in IBS rats [19].
While prior studies have conducted meta-analyses of herbal remedies for IBS [20], there has been no systematic evaluation or meta-analysis of the A-P CHM for IBS to date.Given this context, we undertook a meta-analysis and systematic evaluation using trial sequential analysis (TSA) of the A-P CHM for IBS.Our objective was to scrutinize the therapeutic efficacy of A-P CHM for IBS and to investigate the necessity for new trials targeting specific outcomes, such as the relief of IBS symptoms.

Search Strategy and Eligibility Criteria
The protocol is registered in the International Prospective Register of Systematic Reviews (PROSPERO) (registration number: CRD42023439087).We searched seven databases: PubMed, Excerpta Medica database (Embase), Ovid Medical Literature Analysis and Retrieval System Online (MEDLINE), the Cochrane Library, SinoMed, the Chinese National Knowledge Infrastructure Databases (CNKI), and Wanfang Database, for data up to May 23, 2023.There were no language restrictions.The subject/medical subject headings (MeSH) terms used for the searches were "Atractylodes macrocephala Koidz" OR " Radix Paeoniae Alba" OR "medicine, Chinese traditional" combined with "irritable bowel syndrome".Search strategies are shown in the supplementary files (S1-4 Table ).Irritable bowel syndrome should be diagnosed according to the Rome criteria or the criteria recommended in the guidelines.In randomized controlled trials (RCTs) examining the effect of herbal medicine (including Atractylodes macrocephala and Paeonia lactiflora) therapies in adult patients (≥18 years) with IBS, the control arms were required to receive a placebo, Western medication, or a physician's "usual management".At least one of the following outcomes had to be presented for evaluation by the included trial: relief of global IBS symptoms; adequate relief response rate; the multiple endpoints recommended by the US Food and Drug Administration for IBS; the IBS severity scoring system (IBS-SSS); duration of therapy: ≥ 4 weeks.We used the time point at the end of the treatment.
The search strategy and search terms used are summarized in Tables 1-4.We excluded studies on Chinese medicine combined with other non-Atractylodes macrocephala-Paeonia lactiflora prescriptions, studies where the control group was treated with traditional Chinese medicine, and studies where the full text of the publication could not be obtained.We also excluded studies that were conference articles and not RCTs.

Outcome Assessment
The primary outcomes assessed were the relief of global IBS symptoms, which were defined by the following: (1) the recent consensus that patients self-reported relief of symptoms; (2) an IBS-SSS score 75 points or conditions improved by one level after the last treatment, compared with before treatment; (3) adopting the United States Food and Drug Administration (FDA) criteria for IBS remission response to both abdominal pain and stool consistency; (4) an adequate relief response rate.
The secondary outcomes included the IBS-SSS score and treatment-related adverse events.

Data Abstraction
Data were extracted independently by WJ B and Z W using standardized extraction forms.Data were extracted as intention-to-treat analyses, with dropouts assumed to be treatment failures, whenever trial reporting allowed this, in accordance with the guidance provided by the Cochrane Handbook for Systematic Reviews of Interventions [21].

Risk of Bias Assessment
The Cochrane Risk of Bias (RoB) tool (RoB 2.0) (https://www.riskofbias.info)[22] was used to assess the risk of bias at the individual study level, which was independently done by JW L and Z W. Each domain of a trial was rated with low, high, or some concerns, and an overall RoB was lastly rated for the trial.discrepancy in the RoB assessment was solved by discussion.

Data Analysis
The data were analyzed using the R software (version 4.1.1,The R Core Team, R Foundation for Statistical Computing, Vienna, Austria).Separate analyses were performed to compare the efficacy and safety of the A-P CHM in all studies compared with placebo or other drugs for the treatment of IBS (e.g., improvement in global symptoms of IBS, IBS-SSS, composite FDA endpoint, FDA endpoint abdominal pain relievers, and FDA endpoint intestinal symptom relievers).
For dichotomous variables, relative ratios (RR) and their corresponding 95% confidence intervals (CIs) were calculated; for continuous variables, mean scores and SD were extracted using a standardized mean difference (SMD) with 95% CIs.Data were presented in forest plots.We planned to assess for evidence of publication bias by applying Egger's test (for which p<0.05 is significant to suggest publication bias) to funnel plots of odds ratios (ORs) [23] for 10 studies or more [24].
The number needed to treat (NNT) and the number needed to harm (NNH) were calculated using the formula NNT or NNH = 1/(control event rate -experimental event rate).
The TSA was performed by TSA 0.9.5.10 beta (https://www.ctu.dk/tsa/) to calculate the required information size (RIS).The type I error was allowed to be 0.05, and the type II error was 0.2 when estimating the RIS.The significance boundaries were calculated based on the O'Brien-Fleming alpha-spending method.For continuous data, we estimated the mean difference and variance based on empirical assumptions generated by software.For dichotomous data, we estimated the mean difference and variance based on the incidence of low risk of bias studies.

Certainty Assessment
The GRADEprofiler (gradeworkinggroup.org)was used to evaluate the overall certainty of evidence across RCTs, which was independently done by WJ B and HL D. The quality of evidence was subsequently classified into four categories (high, moderate, low, and very low) according to the corresponding evaluation criteria [26].

Results
We searched 3,117 article citations using the search strategy; of these, 104 published articles appeared to be relevant and were retrieved for further evaluation, of which, for various reasons, 80 were excluded, and finally 24 eligible articles with 3,768 participants met our inclusion criteria (Figure 1).

FIGURE 2: Risk of bias assessment of studies FIGURE 3: The bias assessment of all articles
The detailed characteristics of individual trials are shown in Table 5.

Relief of Global IBS Symptoms
Sixteen trials provided data for the outcome assessment of the proportion of patients who were "relief of global IBS symptoms" responders.Seven trials [27,28,[30][31][32][33][34] were A-P CHM (n=706) vs. placebo (n=710); A-P CHM had significantly a higher proportion of global IBS symptom relief (RR 1.55 (95%CI, 1.21-1.99);I 2 =69%; p < 0.01) NNT=6 (Figure 4a).Based on the visual inspection of the funnel plot, we found no asymmetry (p=0.246, Figure 5).Results from sensitivity analysis showed that the exclusion of any single study did not influence the overall estimate (Figure 6).The evidence relating to A-P CHM and placebo for global IBS symptom relief was downgraded to low quality using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) protocol (Table 6).fewer to 100 more) 1 I 2 = 69%; 2 The confidence interval is wide; 3 Bias risk assessment based on Rob2; 4 The risk of bias in 4 studies is of some concern; 5 Forest map overlap is not good; 6 Results based on the risk of bias; 7 I 2 = 93% 8 Results based on the risk of bias; The confidence interval of forest map is wide; 9 I 2 = 88%; 10 I 2 = 85%; 11 The sample size is small; 12 -0.99 to -0.43; 13 There is some concern about the risk of bias in the 10 studies, high about the risk of bias in the 1 study; 14 -1.09,-0.6;15 There is some concern about the risk of bias in one study The TSA analysis showed that according to the previous meta-analysis [50], which assumed 27% placebo in the proportion of relief of global IBS symptoms and relative risk reduction (RRR)=30%, the RIS required 3361 participants and the accrued sample size (n=1,416) of this meta-analysis.Figure 4b shows that the cumulative Z-curve crosses both the conventional and TSA bounds, indicating that although the cumulative amount of information does not reach the desired value, no more trials are needed to obtain a positive conclusion in advance.

IBS: irritable bowel syndrome; A-P CHM: Atractylodes macrocephala-Paeonia lactiflora class formula
Results from sensitivity analysis showed that the exclusion of any single study did not influence the overall estimate (Figure 9).The evidence relating to A-P CHM and Western medication for global IBS symptom relief was downgraded to low quality using the GRADE protocol (Table 6).The TSA analysis showed that assuming 60% Western medication [31] in the proportion of relief of global IBS symptoms and RRR=20%, the RIS required 1,846 participants and the accrued sample size (n=1586) of this meta-analysis.Figure 7b shows that the cumulative Z-curve crosses both the conventional and TSA bounds, indicating that although the cumulative amount of information does not reach the desired value, no more trials are needed to obtain a positive conclusion in advance.

The FDA Endpoint Responders
Only one trial [31] provided data for the outcome assessment of the proportion of patients who were composite FDA endpoint responders; there were 166 responders out of 348 in A-P CHM, placebo was 47/348, and pinaverium was 141/348.Five trials provided data for the outcome assessment of the proportion of patients who were in the FDAendpoint abdominal pain relief group.Three trials [29,31,35] were A-P CHM (n=497) vs. placebo (n=496); A-P CHM had significantly a higher proportion of FDA endpoint abdominal pain relief (RR 1.94 (95%CI, 1.65-2.27);I 2 =0%; p = 0.82) NNT=3 (Figure 10a).The evidence relating to A-P CHM and the placebo of FDA endpoint abdominal pain relief was of high quality using the GRADE protocol (Table 6).The TSA analysis showed that assuming 34% placebo [50] in the proportion of relief of FDA endpoint abdominal pain and RRR=24%, the RIS required 989 participants and the accrued sample size (n=993) of this meta-analysis.Figure 10b shows that the cumulative Z curve crossed both the conventional and TSA bounds, and the cumulative amount of information reached the desired value, indicating a statistically significant difference between A-P CHM and placebo.In the two trials [31,38], A-P CHM (n=432) vs. pinaverium (n=432), there was no significant difference between the two groups (RR 0.96 (95% CI, 0.69-1.33);I2 =93%; p 0.01) NNT=9 (Figure 10a).The evidence relating to A-P CHM and pinaverium of FDA endpoint abdominal pain relief was downgraded to very low quality using the GRADE protocol (Table 6).
Five trials provided data for the outcome assessment of the proportion of patients who received FDA endpoint stool consistency relief.Three trials [29,31,35] were A-P CHM (n=497) vs. placebo (n=496); A-P CHM had significantly a higher proportion of FDA endpoint stool consistency relief (RR 2.67 (95%CI, 1.64-4.37);I 2 =88%; p <0.01) NNT=2 (Figure 11a).The evidence relating to A-P CHM and placebo of FDA endpoint stool consistency relief was downgraded to low quality using the GRADE protocol (Table 6).The TSA analysis showed that assuming an 18% placebo in the proportion of relief of FDA endpoint stool consistency [50] and RRR=40%, the RIS required 3,445 participants and the accrued sample size (n=993) of this meta-analysis.Figure 11b shows that the cumulative Z curve crossed the traditional boundary value but did not cross the trial sequential boundaries, and the cumulative amount of information has fallen short of expectations, indicating A-P CHM was better than placebo for relief of FDA endpoint stool consistency and possible false positive results.Two trials [50] were A-P CHM (n=432) vs. pinaverium (n=432); A-P CHM had significantly a higher proportion of FDA endpoint stool consistency relief (RR 1.33 (95%CI, 1.06-1.67);I2 =85%; p=0.01]NNT=6 (Figure 11a).The evidence relating to A-P CHM and pinaverium of FDA endpoint stool consistency relief was downgraded to low quality using the GRADE protocol (Table 6).The TSA analysis showed that assuming 51% pinaverium in the proportion of relief of FDA endpoint stool consistency [31] and RRR=25%, the RIS required 3,657 participants and the accrued sample size (n=864) = 864 in this meta-analysis).Figure 11c shows that the cumulative Z curve crossed the traditional boundary value but did not cross the trial sequential boundaries, and the cumulative amount of information has fallen short of expectations, indicating A-P CHM was better than pinaverium in terms of relief of FDA endpoint stool consistency and possible false positive results.

The IBS-SSS Score
Sixteen trials provided data for the outcome assessment of the IBS-SSS score.Fourteen trials [37][38][39][40][41][42][43][44][45][46][47][48][49] were A-P CHM (n=648) vs.Western medication (n=581); A-P CHM IBS-SSS score was lower than that of Western medication (SMD -1.19, 95% CI, -1.61--0.76);I 2 =89%; p <0.01) (Figure 12a).Based on the visual inspection of the funnel plot, we found no asymmetry (p=0.0622, Figure 13).Results from sensitivity analysis showed that the exclusion of any single study did not influence the overall estimate (Figure 14).The evidence relating to A-P CHM and Western medication's IBS-SSS score was downgraded to very low quality using the GRADE protocol (Table 6).The TSA analysis showed that the RIS required 267 participants, and the accrued sample size (n=1229) of this meta-analysis.Figure 12b shows that the cumulative Z curve crossed trial sequential boundaries, and the cumulative amount of information has reached the desired value, indicating a statistically significant difference between A-P CHM and Western medication.
Four trials [10,31,32,39] were A-P CHM (n=597) vs. pinaverium (n=597); there was no significant difference between the two groups (RR 1.05 (95%CI, 0.79-1.40);I2=0%; p=0.67)NNH=81 (Figure 15a).The evidence relating to A-P CHM vs. pinaverium of adverse events was downgraded to low quality using the GRADE protocol (Table 6).The TSA analysis showed that the RIS required 12,865 participants, and the accrued sample size (n=1194) of this meta-analysis.Figure 15b shows that the cumulative Z curve did not cross the trial sequential boundaries, and the cumulative amount of information did not reach the expected value, indicating that further studies are needed for validation.

Discussion
Our meta-analysis incorporated 24 studies.One study by Bensoussan et al. [27] focused on irritable bowel syndrome with constipation (IBS-C), irritable bowel syndrome with diarrhea (IBS-D), and irritable bowel syndrome mixed type (IBS-M); another study by Bensoussan et al. [28] specifically addressed IBS-C, while the remaining primarily catered to IBS-D.This leads us to conjecture that A-P CHM is more beneficial for the diarrhea-predominant type of irritable bowel syndrome.
Regarding the primary outcomes, A-P CHM demonstrated a significantly higher rate of global lBS symptom relief compared to either placebo or Western medication, with NNTs of 6 (vs.placebo) and 14 (vs.Western medication).These findings were further substantiated by TSA analysis; A-P CHM had significantly a higher proportion of FDA endpoint abdominal pain relief than placebo, with an NNT of 3.This was further corroborated by TSA analysis, indicating that our meta-analyses had an adequate sample size.The Atractylodes macrocephala-Paeonia lactiflora class formula had significantly a higher proportion of stool consistency relief than placebo or pinaverium; the NNT was 2 (vs.placebo) and 6 (vs.pinaverium), but the TSA analysis showed these possible false-positive results.
In relation to secondary outcomes, the lBS-SSS score for A-P CHM was lower than that for either Western medication or placebo at the conclusion of the treatment, a finding that was further validated by TSA analysis and confirmed the sufficient sample size of our meta-analysis.We also found that for adverse events, there was no significant difference between A-P CHM and placebo or Western medication; the NNH was 33 (vs.placebo) and 81 (vs.pinaverium); this was further confirmed by the TSA analysis.
Additionally, we utilized the GRADEprofiler to assess the overall certainty of evidence across RCTs.The evidence for FDA endpoint abdominal pain relief of A-P CHM and placebo was of high quality, while the evidence for IBS-SSS score and adverse events was of moderate quality.Evidence for other measures was downgraded to low or very low, indicating an overall low quality [51].It was not until 2012 that the U.S. Food and Drug Administration recommended the use of endpoints assessing improvement in abdominal pain, improvement in bowel habits, or both in IBS-D.Since the publication of these recommendations, many RCTs of tightly designed drugs have followed these endpoints [52][53][54], although some have come close to these endpoints [55,56].However, to the best of our knowledge, meta-analyses of herbal medicines for IBS have not used this endpoint.This information is important because it can inform sample size calculations for future high-quality RCTs.In addition, because the magnitude of the placebo response may affect the likelihood that a drug will show significantly better efficacy than a placebo, the choice of endpoints used to confirm efficacy may determine the success or failure of a drug.Consequently, we conducted a systematic review and meta-analysis of the efficacy of A-P CHM in treating irritable bowel syndrome, using the FDArecommended endpoints as one of the outcome measures.
With 22 studies targeting IBS-D, it is reasonable to infer that A-P CHM may be more effective for diarrheapredominant irritable bowel syndrome.According to Chinese medicine theory, IBS-D, despite its location in the intestines, is closely associated with the liver and rooted in the spleen.The liver is associated with the wood element, and the spleen is linked to the earth element.The record Jingyue quanquan notes: "Encountering anger leads to diarrhea...This involves diseases of both the liver and spleen, where the liver counters the earth element and the spleen receives Qi."In Medical Examination Volume II, it is stated that "The spleen is responsible for diarrhea, while the liver is responsible for pain.The liver deals with reality, and the spleen handles the virtual.Spleen deficiency leads to a solid liver, resulting in pain and diarrhea."Therefore, "pain" is primarily attributed to excessive wood strength, and "diarrhea" is linked to spleen deficiency and dampness.
In the context of diarrhea-type irritable bowel syndrome, the primary mechanism involves deficiencies in the liver and spleen, with wood overpowering the earth.The treatment approach involves addressing the imbalance by tonifying the earth element to counteract the wood element's excess [57].
Atractylodes macrocephala, with its bitter and sweet qualities, is warming and associated with the spleen and stomach meridians.It strengthens the spleen and alleviates diarrhea.Paeonia lactiflora, bitter and sour, slightly cold, pertains to the liver and spleen meridian.It softens the liver and relieves pain.The Chinese herbal formula, primarily composed of Atractylodes macrocephala and Paeonia lactiflora, aims to achieve therapeutic effects by addressing the interaction of wood and earth in the context of "diarrhea of wood in the earth."To the best of our knowledge, this study is the first to investigate the effectiveness of A-P CHM in the treatment of IBS.
The limitations of this systematic review and meta-analysis come from the studies available for synthesis.Nearly half of the trials were at high risk of bias or had some concerns, especially in trials published in Chinese journals, and the most common were deviations from intended interventions, especially the implementation of blinding, which suggests that attention should be paid to the design of blinding in future related studies to ensure the quality of the trials.And there is evidence of heterogeneity between the RCTs of A-P CHM and placebo and Western medicine treatments, and there is also evidence of publication bias between A-P CHM and Western medicines in improving the overall symptoms of IBS.This may lead to an overestimation of the treatment effect.

Conclusions
Our study demonstrated the efficacy of A-P CHM in alleviating global IBS symptoms, improving FDA endpoint stool consistency, and reducing IBS-SSS scores.The findings of our meta-analysis were corroborated by the TSA analysis.Concurrently, no significant difference was observed in adverse events when compared with placebo or pinaverium, suggesting that A-P CHM might serve as a potential treatment for IBS patients, particularly those with IBS-D.This might provide a theoretical foundation for future optimization of herbal formulas for IBS-D.The overall certainty of the evidence was not high, underscoring the need for more rigorous RCTs in the future.

FIGURE 1 :
FIGURE 1: A flowchart showcasing the selection of studies Embase: Excerpta Medica database; MEDLINE: Medical Literature Analysis and Retrieval System Online; RCT: randomized controlled trial

FIGURE 4 :
FIGURE 4: Meta-analysis and TSA analysis of A-P CHM vs. placebo in the proportion of patients who were "relief of global IBS symptoms" responders 4a: The results of the meta-analysis; 4b: The results of TSA: error α=5%, β=20%, incidence in the control arm =27%, RRR=30%.The red vertical line represented the RIS for the comparison, the blue line represented the cumulative Z curve, and the black dots represented the included studies.TSA: trial sequential analysis; A-P CHM: Atractylodes macrocephala-Paeonia lactiflora class formula; IBS: irritable bowel syndrome; RRR: relative risk reduction; RIS: required information size

FIGURE 7 :
FIGURE 7: Meta-analysis and TSA analysis of A-P CHM vs.Western medication in the proportion of patients who were "relief of global IBS symptoms" responders 7a: The results of the meta-analysis; 7b: The results of TSA: error α=5%, β=20%, incidence in the control arm =60%, RRR = 20%.The red vertical line represents the RIS for the comparison, the blue line represents the cumulative Z curve, and the black dots represent the included studies.TSA: trial sequential analysis; A-P CHM: Atractylodes macrocephala-Paeonia lactiflora class formula; IBS: irritable bowel syndrome; RRR: relative risk reduction; RIS: required information size

FIGURE 8 :
FIGURE 8: The funnel plot of global IBS symptom relief (A-P CHM vs.

FIGURE 9 :
FIGURE 9: The sensitivity analysis of global IBS symptom relief (A-P CHM vs.Western medication) IBS: irritable bowel syndrome; A-P CHM: Atractylodes macrocephala-Paeonia lactiflora class formula

FIGURE 12 :
FIGURE 12: Meta-analysis and TSA analysis of the IBS-SSS score 12a: The results of the meta-analysis of A-P CHM vs.Western medication in the IBS-SSS score; 12b: The results of TSA of A-P CHM vs.Western medication in the IBS-SSS score: error α=5%, β=20%; 12c: The results of the meta-analysis of A-P CHM vs. placebo in the IBS-SSS score.The red vertical line represents the RIS for the comparison, the blue line represents the cumulative Z curve, and the black dots represent the included studies.TSA: trial sequential analysis; A-P CHM: Atractylodes macrocephala-Paeonia lactiflora class formula; IBS-SSS: irritable bowel syndrome-symptom severity score; RIS: required information size

TABLE 6 : The overall certainty of evidence across RCTs CHM
: Chinese herbal medicine; RCT: randomized controlled trial; IBS: irritable bowel syndrome; IBS-SSS: IBS severity scoring system; AR: adequate relief; RR: relative ratio; AE: adverse event; RoB: risk of bias