The efficacy and safety of Cangfu Daotan decoction in the treatment of infertility associated with polycystic ovary syndrome: a systematic review and meta-analysis.

A total of 644 records were identified through database searches, including 622 from Chinese databases (CNKI, Wanfang, VIP) and 22 from international databases (PubMed, Embase, Cochrane Library, Web of Science). Additional records were retrieved from clinical trial registries (ChiCTR). After automated deduplication in EndNote 21, 312 duplicates were removed, leaving 331 unique records. Title and abstract screening excluded 172 studies that did not meet eligibility criteria. Full-text review of 159 articles against PICOS standards led to the exclusion of 140 studies, for the following reasons: irrelevant outcome measures ( n  = 84), combined use of other TCM therapies ( n  = 51), small sample size (< 50, n  = 2), non-matching interventions ( n  = 2), and incompatible study populations ( n  = 1). One additional study was excluded due to unavailable data despite the author’s correspondence. Ultimately, 19 RCTs met the inclusion criteria and were included in the final analysis. The study selection process is shown in the PRISMA flow diagram (Fig.  1 ). Fig. 1 Flow diagram of the study selection process. Abbreviations: CNKI, China National Knowledge Infrastructure; VIP, China Science and Technology Journal Database; ChiCTR, Chinese Clinical Trial Registry; TCM, Traditional Chinese Medicine Flow diagram of the study selection process. Abbreviations: CNKI, China National Knowledge Infrastructure; VIP, China Science and Technology Journal Database; ChiCTR, Chinese Clinical Trial Registry; TCM, Traditional Chinese Medicine Nineteen RCTs [ 17 – 34 ], all conducted in China and published between 2013 and 2025, were included, involving a total of 2,181 women with PCOS (1,108 in intervention groups and 1,073 in control groups). Diagnostic criteria varied: seven studies [ 12 , 17 – 21 , 33 ] used World Health Organization (WHO) criteria, ten [ 22 , 24 – 27 , 29 – 31 , 34 ] applied Chinese guidelines, and two [ 23 , 32 ] did not specify diagnostic criteria. Treatment duration ranged from three to four menstrual cycles. Four trials [ 22 , 26 – 28 ] employed the classical CFDTT formula, while 15 [ 17 – 21 , 23 – 25 , 29 – 34 ] used modified formulations tailored to patients’ concomitant symptoms or comorbidities. In 18 studies[ 31 – 34 ], , CFDTT (original or modified) was combined with conventional pharmacotherapy, with the control groups receiving conventional pharmacotherapy alone. One trial [ 30 ] tested CFDTT monotherapy against conventional pharmacotherapy. Detailed characteristics of the included studies are summarized in Table 1 . Table 1 Characteristics of the included studies Author/ Year Sample size Age (years) Course of disease Duration intervention Outcome index Chen, et al. 2015 [ 17 ] 45/45 E: 34.45 ± 6.23 C: 35.86 ± 5.85 E: 7.48 ± 0.89years C: 7.27 ± 0.86years 3months E: CFDTT+Diane-35 + CC C: Diane-35 + CC ①②④⑤ Chen 2020 [ 18 ] 33/33 E:26.82 ± 1.56 C:26.80 ± 1.54 E:2.45 ± 0.89 years C:2.44 ± 0.88 years 4months E: CFDTT+Diane-35 + Met C: Diane-35 + Met ①② Ding, et al. 2013 [ 20 ] 50/50 E: 27.2 ± 4.8 C: 25.9 ± 5.7 1-6years 3months E: CFDTT + CC+HMG + HCG C: CC + HMG+HCG ②⑥ Ding, et al. 2014 [ 19 ] 195/160 28. 5 ± 11. 5 7.6 ± 3.7 years 3months E: CFDTT + CC+HMG + HCG C: CC + HMG+HCG ②⑥ Du, et al. 2018 [ 29 ] 50/50 E: 25.4 ± 3.2 C: 26.1 ± 3.0 NR 3months E: CFDTT+Diane-35 C: Diane-35 ①② Huang, et al. 2021 [ 25 ] 100/100 E:25.80 ± 3.85 C: 25.68 ± 3.62 E:1.57 ± 0.32 years C: 1. 62 ± 0.34 years 3months E: CFDTT+Diane-35 + Met + CC+HCG+Prog C: Diane-35 + Met+Prog ①② Huang 2018 [ 21 ] 42/42 E: 24.87 ± 4.67 C: 24.59 ± 4.92 E: 3.35 ± 0.93 years C: 3.25 ± 0.86 years 3months E: CFDTT + Met C: Met ①②③ Jin 2020 [ 23 ] 45/45 E: 25.89 ± 7.09 C: 25.93 ± 7.12 E: 5.24 ± 1.89 years C: 5.01 ± 1.56 years 4months E: CFDTT+Diane-35 + CC C: Diane-35 + CC ①② Lan, et al. 2019 [ 34 ] 45/46 E: 35.82 ± 3.83 C: 36.26 ± 3.63 E: 2.01 ± 0.49 years C: 1.87 ± 0.45 years 3months E: CFDTT + CC+Met + HCG C: CC + Met+HCG ①②⑥ Li, et al. 2017 [ 30 ] 58/58 E: 27.56 ± 3.25 C: 27.90 ± 3.31 E: 3.15 ± 0.36 years C: 3.24 ± 0.51 years 3months E: CFDTT + CC+Diane-35 + HMG + HCG C: CC+Diane-35 + HMG + HCG ①②⑥ Li, et al. 2018 [ 24 ] 124/124 E: 25.64 ± 1.30 C: 26.17 ± 1.82 E: 4.35 ± 2.41 years C: 4.52 ± 1.94 years 3months E: CFDTT + LET+HMG + HCG+Prog C: LET + HMG+HCG+Prog ②③⑤ Li, et al. 2023 [ 22 ] 36/36 E: 27.44 ± 4.77 C: 27.64 ± 5.01 E: 2.89 ± 1.35 years C: 2.92 ± 1.08 years 3months E: CFDTT + Met C: Met ①②④ Liu 2023 [ 35 ] 36/35 E: 31.00 ± 1.29 C: 30.43 ± 1.31 E: 3.28 ± 1.39 years C: 3.31 ± 1.37 years 3months E: CFDTT+Diane-35 + HCG C: Diane-35 + HCG ① Lu, et al. 2020 [ 26 ] 30/30 E: 28.70 ± 3.82 C: 28.70 ± 3.82 NR 3months E: CFDTT + CC C: CC ①②⑤⑥ Luo 2024 [ 28 ] 37/37 E: 28.22 ± 3.78 C: 27.95 ± 3.69 E: 1.4 ± 3.2 years C: 1.5 ± 3.1 years 4months E: CFDTT + CC+lifestyle intervention C: CC+lifestyle intervention ①②③④ Zeng 2017 [ 32 ] 60/60 E: 32.35 ± 5.64 C: 32.23 ± 5.60 E: 3.56 ± 1.02 years C: 3.61 ± 1.05 years 4months E: CFDTT+Diane-35 + Met C: Diane-35 + Met ①②⑥ Zhang, et al., 2022 [ 31 ] 36/36 E: 30.29 ± 4.41 C: 30.46 ± 4.17 E: 6.02 ± 1.56 years C: 6.51 ± 1.83 years 3months E: CFDTT C: CC ②④⑥ Zhang 2018 [ 33 ] 56/56 E: 29.74 ± 2.03 C: 29.62 ± 1.90 E: 4.25 ± 1.03 years C: 4.18 ± 1.01 years 3months E: CFDTT + CC C: CC ①②④ Zhang 2023 [ 36 ] 30/30 E: 23.45 ± 4.32 C: 23.36 ± 4.47 E: 2.46 ± 1.42 years C: 2.23 ± 1.25 years 3months E: CFDTT + LET C: LET ①②③ NR Not reported, E Experimental group, C Control group, Intervention, CC Clomifene Citrate Capsules， Diane-35 Ethinylestradiol and Cyproterone Acetate Tablets， Met Metformin Hydrochloride Tablets， Prog Progesterone， HCG Human Chorionic Gonadotropin， HMG Human Menopausal Gonadotropin Outcome index:①Ovulation rate; ②Pregnancy rate; ③Incidence of adverse events; ④Ovarian volume; ⑤Dominant follicle count; ⑥Endometrial thickness Characteristics of the included studies Chen, et al. 2015 [ 17 ] E: 34.45 ± 6.23 C: 35.86 ± 5.85 E: 7.48 ± 0.89years C: 7.27 ± 0.86years E: CFDTT+Diane-35 + CC C: Diane-35 + CC E:2.45 ± 0.89 years C:2.44 ± 0.88 years E: CFDTT+Diane-35 + Met C: Diane-35 + Met E: 27.2 ± 4.8 C: 25.9 ± 5.7 E: CFDTT + CC+HMG + HCG C: CC + HMG+HCG E: CFDTT + CC+HMG + HCG C: CC + HMG+HCG E: 25.4 ± 3.2 C: 26.1 ± 3.0 E: CFDTT+Diane-35 C: Diane-35 E:25.80 ± 3.85 C: 25.68 ± 3.62 E:1.57 ± 0.32 years C: 1. 62 ± 0.34 years E: CFDTT+Diane-35 + Met + CC+HCG+Prog C: Diane-35 + Met+Prog E: 24.87 ± 4.67 C: 24.59 ± 4.92 E: 3.35 ± 0.93 years C: 3.25 ± 0.86 years E: CFDTT + Met C: Met E: 25.89 ± 7.09 C: 25.93 ± 7.12 E: 5.24 ± 1.89 years C: 5.01 ± 1.56 years E: CFDTT+Diane-35 + CC C: Diane-35 + CC E: 35.82 ± 3.83 C: 36.26 ± 3.63 E: 2.01 ± 0.49 years C: 1.87 ± 0.45 years E: CFDTT + CC+Met + HCG C: CC + Met+HCG E: 27.56 ± 3.25 C: 27.90 ± 3.31 E: 3.15 ± 0.36 years C: 3.24 ± 0.51 years E: CFDTT + CC+Diane-35 + HMG + HCG C: CC+Diane-35 + HMG + HCG E: 25.64 ± 1.30 C: 26.17 ± 1.82 E: 4.35 ± 2.41 years C: 4.52 ± 1.94 years E: CFDTT + LET+HMG + HCG+Prog C: LET + HMG+HCG+Prog E: 27.44 ± 4.77 C: 27.64 ± 5.01 E: 2.89 ± 1.35 years C: 2.92 ± 1.08 years E: CFDTT + Met C: Met E: 31.00 ± 1.29 C: 30.43 ± 1.31 E: 3.28 ± 1.39 years C: 3.31 ± 1.37 years E: CFDTT+Diane-35 + HCG C: Diane-35 + HCG E: 28.70 ± 3.82 C: 28.70 ± 3.82 E: CFDTT + CC C: CC E: 28.22 ± 3.78 C: 27.95 ± 3.69 E: 1.4 ± 3.2 years C: 1.5 ± 3.1 years E: CFDTT + CC+lifestyle intervention C: CC+lifestyle intervention E: 32.35 ± 5.64 C: 32.23 ± 5.60 E: 3.56 ± 1.02 years C: 3.61 ± 1.05 years E: CFDTT+Diane-35 + Met C: Diane-35 + Met E: 30.29 ± 4.41 C: 30.46 ± 4.17 E: 6.02 ± 1.56 years C: 6.51 ± 1.83 years E: CFDTT C: CC E: 29.74 ± 2.03 C: 29.62 ± 1.90 E: 4.25 ± 1.03 years C: 4.18 ± 1.01 years E: CFDTT + CC C: CC E: 23.45 ± 4.32 C: 23.36 ± 4.47 E: 2.46 ± 1.42 years C: 2.23 ± 1.25 years E: CFDTT + LET C: LET NR Not reported, E Experimental group, C Control group, Intervention, CC Clomifene Citrate Capsules， Diane-35 Ethinylestradiol and Cyproterone Acetate Tablets， Met Metformin Hydrochloride Tablets， Prog Progesterone， HCG Human Chorionic Gonadotropin， HMG Human Menopausal Gonadotropin Outcome index:①Ovulation rate; ②Pregnancy rate; ③Incidence of adverse events; ④Ovarian volume; ⑤Dominant follicle count; ⑥Endometrial thickness The methodological quality of the included studies was assessed using the Cochrane ROB-2 tool across five domains. Randomization process: Concerns were raised in 100% (19/19) of the studies, as none reported allocation concealment, and several failed to detail random sequence generation methods. Intervention deviation: 53% (10/19) of the studies raised some concerns, while 46% (9/19) were judged to have a high risk of bias. Missing outcome data: 100% (19/19) of the studies were classified as low risk, indicating that incomplete data were appropriately handled and unlikely to introduce substantial bias. Outcome measurement: 11% (2/19) of studies were rated as low risk, while 89% (17/19) raised some concerns, primarily due to the lack of blinding of outcome assessors. Outcome selection: 95% (18/19) of studies were classified as having some concerns, and 5% (1/19) as low risk, suggesting potential selective reporting of outcomes, However, no studies in this domain were rated as having a high risk of bias. Overall, approximately 50% of the studies were rated as having a high risk of bias, with the remaining 50% raising some concerns. These findings highlight substantial methodological heterogeneity across the included trials, particularly in terms of blinding practices, compliance monitoring, and reporting standards. The assessment results are presented in Figure 2 . Fig. 2 Risk of bias assessment for included studies. A Risk of bias graph; ( B ) Risk of bias summary Risk of bias assessment for included studies. A Risk of bias graph; ( B ) Risk of bias summary Fifteen trials ( n = 1,406) reported ovulation outcomes [ 17 , 18 , 21 – 23 , 31 – 34 ]. Meta-analysis using a fixed-effect model showed that CFDTT combined with conventional conventional pharmacotherapy may improve ovulation rates compared with conventional conventional pharmacotherapy alone (RR = 1.38, 95% CI: 1.29–1.49; p < 0.00001; I² =0%, p = 0.61; Fig. 3 A). Subgroup analyses stratified by age ( p = 0.92), disease duration ( p = 0.94), and intervention modality ( p = 0.93) revealed no significant effect modification. Sensitivity analysis confirmed the robustness of results (Supplementary S6A). Fig. 3 Forest plot for primary outcomes. A Ovulation rate ( B ) Pregnancy rate ( C ) Incidence of adverse event Forest plot for primary outcomes. A Ovulation rate ( B ) Pregnancy rate ( C ) Incidence of adverse event Seventeen trials ( n = 2,071) reported pregnancy rates [ 17 – 26 , 31 – 34 ]. Pooled analysis under a fixed-effect model indicated that combined therapy increased pregnancy rates compared with monotherapy (RR = 1.58, 95% CI: 1.43–1.75; p < 0.00001; I² =10%, p = 0.33; Fig. 3 B). Subgroup analyses by age ( p = 0.31), disease duration ( p = 0.33), and intervention modality ( p = 0.21) showed consistent benefits across strata. Sensitivity analyses yielded stable results (Supplementary S6B). One trial ( n = 72) compared CFDTT monotherapy with conventional pharmacotherapy [ 30 ]. Although pregnancy rates were higher with CFDTT, the difference was not statistically significant (RR = 1.07, 95% CI: 0.61–1.88; p = 0.81; Fig. 3 C), likely reflecting insufficient statistical power. Among the 19 included studies, five reported adverse events. One study reported no adverse reactions during treatment, two studies documented mild gastrointestinal effects, and two focused on reproductive system–related events (Table  2 ). Huang 2018 observed 4 cases (9.52%) of adverse reactions in the combination therapy group, including 2 cases of nausea and vomiting and 2 cases of intermittent diarrhea, compared with 2 cases (4.76%) of nausea and vomiting in the control group. Luo 2024 reported 5 cases (13.51%) in the combination group (1 nausea and vomiting, 3 headaches, 1 bone pain) versus 4 cases (10.81%) in the control group (1 nausea and vomiting, 2 headaches, 1 bone pain). Li et al. 2018 found 12 cases (9.67%) of luteinized unruptured follicle syndrome (LUFS) and 1 case (0.81%) of ovarian hyperstimulation syndrome (OHSS) in the combination group, compared with 34 cases (27.42%) of LUFS and 1 case (4.84%) of OHSS in the control group. Zhang 2023 reported 1 case (10.0%) of OHSS in the combination group and 6 cases (10.0%) in the control group. Overall, adverse events were generally mild, with gastrointestinal and reproductive system effects being the most commonly reported. Combination therapy did not appear to increase the overall incidence of adverse reactions compared with conventional pharmacotherapy. The adverse events are presented in Table  2 . Table 2 Adverse events Author Year Groups Digestive system Nervous system Reproductive System Other Nausea and vomiting Intermittent diarrhea Headache LUFS OHSS Bone pain Huang 2018 [ 21 ] E 2 2 C 2 Luo 2024 [ 28 ] E 1 3 1 C 1 2 1 Li, et al. 2018 [ 24 ] E 12 1 C 34 1 Zhang 2023 [ 36 ] E 1 C 6 Adverse events Four trials ( n = 466) reported adverse event [ 21 , 24 , 28 , 34 ]. A random-effects model was applied due to moderate heterogeneity ( I² =62%, p = 0.05). Incidence of adverse reactions did not differ significantly between groups (RR = 0.60, 95% CI: 0.22–1.68; p = 0.33; Fig. 3 D). Sensitivity analysis confirmed the stability of results (Supplementary S6C). Four trials ( n = 348) [ 17 , 22 , 28 , 32 ] evaluated ovarian volume changes in PCOS patients. Due to considerable heterogeneity ( I² =94%, p < 0.0001), a random-effects model was applied. The combination of CFDTT with conventional pharmacotherapy was associated with greater ovarian volume reduction compared with conventional pharmacotherapy alone (SMD = 2.14, 95% CI: 1.08–3.21; p < 0.00001; Fig. 4 A), indicating meaningful morphological improvement. Sensitivity analysis confirmed the robustness of this finding (Supplementary Material S6D). Three trials ( n = 398) [ 17 , 24 , 26 ] assessed dominant follicle development in PCOS patients. Substantial heterogeneity ( I² =78%, p = 0.01) prompted the use of a random-effects model. Patients receiving CFDTT alongside conventional pharmacotherapy had higher dominant follicle counts than those receiving pharmacotherapy alone (SMD = 1.47, 95% CI: 0.93–2.00; p = 0.01; Fig. 4 B). This effect is consistent with folliculogenesis-enhancing mechanisms reported in gonadotropin studies. Sensitivity analysis confirmed methodological stability (Supplementary Material S6E), though residual heterogeneity may reflect differences in ultrasonographic assessment protocols. Three trials ( n = 327) [ 12 , 31 , 33 ] evaluated endometrial thickness. High heterogeneity ( I² =82%, p = 0.004) necessitated a random-effects model. CFDTT combined with conventional pharmacotherapy improved endometrial thickness compared with monotherapy (SMD=-1.95, 95% CI: -2.58 to -1.32; p < 0.00001; Fig. 4 C). Sensitivity analysis supported the reliability of these results (Supplementary Material S6F). Fig. 4 Forest plot for primary outcomes. A Ovarian volume ( B ) Dominant follicles count ( C ) Endometrial thickness Forest plot for primary outcomes. A Ovarian volume ( B ) Dominant follicles count ( C ) Endometrial thickness One trial ( n = 72) [ 30 ] compared CFDTT monotherapy with conventional pharmacotherapy and observed a non-significant reduction in endometrial thickness (SMD=-0.30, 95% CI: -0.61 to -0.01; p = 0.06; Fig. 4 D). This limited effect may reflect phenotypic heterogeneity in PCOS populations, particularly among patients with iatrogenic thin endometrium following combined oral contraceptive (COC) pretreatment. Approximately 17.9% of COC-treated PCOS women exhibit endometrial thickness < 7 mm during ovulation induction [ 34 ], a subgroup potentially less responsive to CFDTT-mediated endometrial modulation. Publication bias was assessed for analyses involving ≥ 10 studies using funnel plots and Egger’s test. For ovulation rates, the funnel plot exhibited asymmetry among study points (Fig.  5 A), with Egger’s test indicating statistical significance ( p  = 0.017), suggestive of potential publication bias. Subsequent trim-and-fill correction (after two iterations) identified no additional studies (Supplementary Material S8A), demonstrating no evidence of asymmetry attributable to publication bias and implying alternative sources for the observed asymmetry. Consistency between initial and corrected analyses confirmed result robustness for this outcome. In contrast, the pregnancy rate funnel plot displayed mild asymmetry (Fig.  5 B), with significant Egger’s test results ( p  = 0.001), indicating possible publication bias.1 Trim-and-fill adjustment (after four iterations) incorporated six additional studies (Supplementary Material S8B), yielding 23 studies with no residual publication bias and a pooled effect size of 1.427 [ 34 ]. Fig. 5 Funnel plots for assessing publication bias. A Ovulation rates; ( B ) Pregnancy rates Funnel plots for assessing publication bias. A Ovulation rates; ( B ) Pregnancy rates The quality of evidence was appraised using the GRADEpro tool across five domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias. Overall, evidence for all outcomes was rated as low to very low (Supplementary Material S9), mainly due to a lack of blinding in most studies, substantial heterogeneity ( I² >50%), and wide confidence intervals. These limitations highlight the need for future multicenter, double-blind, placebo-controlled RCTs with standardized protocols to definitively determine the efficacy and safety of CFDTT in managing PCOS.

This systematic review and meta-analysis followed the Cochrane Handbook for Systematic Reviews of Interventions (Version 6.3, 2022) and adhered to the PRISMA 2020 guidelines [ 13 , 14 ] (Supplementary Material S1). The protocol was prospectively registered in PROSPERO (CRD420251066558) before data extraction. A systematic literature search was conducted in electronic databases from inception to March 2025. The searched databases included PubMed, the Cochrane Library, Embase, Web of Science Core Collection, China National Knowledge Infrastructure (CNKI), Wanfang Data Knowledge Service Platform (Wanfang), and the China Science and Technology Journal Database (VIP). Furthermore, ongoing and unpublished trials were searched through ClinicalTrials.gov and the Chinese Clinical Trials Registry (ChiCTR). The search combined subject headings and free-text terms such as: “Cangfu Daotan Decoction”, “Cangfu Daotan Pill”, “Cangfu Daotan”, “cangfudaotan”, “Polycystic Ovary Syndrome”, “PCOS”, “Ovarian Syndrome, Polycystic”, “Sclerocystic Ovarian Degeneration”, “Stein-Leventhal Syndrome”, and “Sclerocystic Ovary”. The detailed search strategies are provided in Supplementary Material S2. Additionally, a manual search and evaluation of the gray literature was also performed. Study design : Only randomized controlled trials (RCTs) published in English or Chinese were included. Participants : Women diagnosed with infertility secondary to PCOS, with no restrictions on age, ethnicity, or sociodemographic characteristics. ‌ Intervention : The experimental group received CFDTT (original or modified formula), in any dosage form (e.g., decoction, granules), alone or combined with conventional pharmacotherapy. The control group received conventional pharmacotherapy or placebo. If both groups used conventional pharmacotherapy, the regimen had to be identical between groups. Comparators : Control groups received either conventional pharmacotherapy alone or a placebo. Study design : Non-randomized controlled trials (e.g., case-control studies, case reports, cross-sectional studies, cohort studies, reviews, expert opinions) were excluded. Trials without accessible outcome data, even after author contact, were excluded. In cases of duplicate publications, the most complete dataset was retained. Participants : Exclusion criteria included women with infertility due to tubal factors, structural reproductive anomalies, endometriosis, ovarian tumors, or other organic lesions; those with endocrine disorders (e.g., Cushing’s syndrome, thyroid disease, hypothalamic dysfunction); patients with severe systemic diseases (hepatic, renal, cardiac, pulmonary, hematologic, or gastrointestinal disorders); and individuals with malignancies or psychiatric illness. Intervention : Trials using multiple concurrent interventions where CFDTT was not the main therapy were excluded, including combinations with other TCM formulations, acupuncture, moxibustion, massage, or acupoint injections. Studies with control groups assigned to non- conventional pharmacotherapy or non-placebo treatments were also excluded. Ovulation rate, clinical pregnancy rate, and incidence of adverse events. Ovarian volume, dominant follicle count, and endometrial thickness. Two reviewers (Yi Wang and Yumei Tang) independently conducted database searches following the predefined strategy. All retrieved records were imported into EndNote 21 for reference management, and duplicates were removed. Titles and abstracts were screened against eligibility criteria, followed by full-text assessment of potentially relevant studies. Selection criteria considered study design, participant characteristics, intervention and control protocols, and outcome measures. Discrepancies were resolved through discussion, with unresolved cases adjudicated by a third reviewer (Jili Xu). Two reviewers (Yi Wang and Yumei Tang) independently extracted data using a standardized electronic form. Extracted information included: study characteristics (author, year, journal, country, language), participant data (sample size, age, disease duration), and methodological details (study design, diagnostic criteria, intervention protocols). All primary and secondary outcomes, as well as reported adverse events, were recorded. Continuous variables reported as medians (with interquartile range or range) were converted to mean ± SD using established algorithms. Missing data were sought from the study authors. Any disagreements were resolved through discussion with a third reviewer (Jili Xu), and the final dataset was verified before entry into the analysis database. Two reviewers (Yi Wang and Yumei Tang) independently evaluated study quality using the Cochrane Risk of Bias 2 (ROB-2) tool [ 15 ], covering: randomization process, deviations from intended interventions, missing outcome data, outcome measurement, and selective reporting. Each domain was rated as “low risk,” “some concerns,” or “high risk.” Discrepancies were resolved through discussion with a third reviewer (Jili Xu). Data were analyzed using Review Manager 5 and Stata 18. Dichotomous outcomes were expressed as risk ratios (RR) with 95% confidence intervals (CIs). Continuous outcomes were reported as standardized mean differences (SMDs) with 95% CIs to account for variations in measurement scales. Heterogeneity was assessed using χ² tests ( p < 0.10 indicating significance) and quantified with the I² statistic. Fixed-effects models were applied when I² < 50% and p ≥ 0.10, otherwise, random-effects models were used, with subgroup analyses to explore heterogeneity. Sensitivity analyses were performed for primary outcomes. Publication bias was assessed using funnel plots (for ≥ 10 studies) and Egger’s regression test ( p < 0.05 considered significant). Where asymmetry was detected, the trim-and-fill method was applied. The quality of evidence for each outcome was assessed using the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) approach [ 16 ]. This framework classifies the certainty of evidence as high, moderate, low, or very low, based on factors such as risk of bias, inconsistency, indirectness, imprecision, and publication bias. To explore potential effect modifiers, subgroup analyses were conducted based on participant age (< 30 vs. ≥30 years), disease duration (< 3 years, ≥ 3 years, or unspecified), and therapeutic modality (endocrine regulation, ovulation induction, or combined therapy). This stratification allowed investigation of heterogeneity across clinically relevant dimensions.

In summary, this study suggests that CFDTT, when used as an adjunct to conventional conventional pharmacotherapy, may provide additional benefits in improving ovulation rates, pregnancy rates, and enhancing endometrial receptivity. However, the evidence supporting the efficacy of CFDTT as a monotherapy is limited, and the question of whether CFDTT alone can serve as a replacement for conventional pharmacotherapy remains uncertain and inadequately addressed by the current literature. Although reported safety data did not indicate increased risk, 73.7% of included studies did not report adverse events, and no studies monitored organ toxicity, so safety conclusions should be interpreted cautiously. The limited number of studies, small sample sizes, methodological shortcomings, and low-quality evidence—particularly inadequate reporting of blinding and allocation concealment—affect the reliability and robustness of results, rendering the evidence inconclusive. Clinically, treatment of PCOS-related infertility should consider each patient’s overall condition to develop individualized strategies. Future research should prioritize high-quality, large-sample, multi-center, randomized, double-blind, placebo-controlled trials with rigorous design standards to provide more reliable evidence for CFDTT’s clinical application. Additionally, the mechanisms by which CFDTT exerts its effects remain incompletely understood, and future studies employing multi-omics approaches could clarify its actions at the genetic, transcriptional, and metabolic levels.

With the ongoing advancement and widespread implementation of evidence-based medicine, systematic reviews and meta-analyses have become indispensable methodologies for evaluating clinical evidence. These studies are regarded as high-quality sources of medical evidence, offering valuable insights for clinical decision-making. The present study aims to assess the efficacy and safety of CFDTT as an adjunctive treatment for infertility associated with PCOS, thereby providing updated evidence for its clinical application. A comprehensive search of both Chinese and English databases was conducted, with rigorous screening procedures, ultimately including 19 randomized controlled trials (RCTs). Meta-analyses were performed on key outcome measures, including ovulation rate, pregnancy rate, incidence of adverse events, ovarian volume, number of dominant follicles, and endometrial thickness. Regarding primary outcomes, CFDTT in combination with conventional pharmacotherapy demonstrated a significantly higher ovulation rate (RR = 1.38, 95% CI [1.29, 1.49]) and pregnancy rate (RR = 1.58, 95% CI [1.43, 1.75]) when compared to conventional pharmacotherapy alone, without a significant increase in the overall risk of adverse events (RR = 0.60, 95% CI [0.22, 1.68]). These results suggest preliminary support for the clinical use of CFDTT, indicating its potential to act synergistically when combined with conventional treatments. However, interpretation of these primary outcomes should be undertaken with caution, given the limitations of the existing evidence. Although the pooled effect sizes for ovulation and pregnancy rates were statistically significant, the quality of evidence, as assessed by the GRADE approach, was rated as low to very low. This is primarily due to the high risk of bias across the included studies, especially in areas such as blinding and allocation concealment, which could have resulted in an overestimation of the true effect of the intervention. Moreover, direct evidence on the efficacy of CFDTT as monotherapy is scarce, with only one study directly comparing CFDTT to conventional pharmacotherapy, finding no significant differences in pregnancy rates or endometrial thickness. Consequently, any assertion that CFDTT is equivalent to conventional pharmacotherapy as a standalone treatment remains premature and warrants validation through additional high-quality, head-to-head studies. Regarding safety outcomes, this study found no significant increase in adverse event risk with the combination of CFDTT and conventional pharmacotherapy (RR = 0.60, 95% CI [0.22, 1.68]). Some studies even suggested that the combination therapy might reduce the incidence of LUFS or OHSS. However, this apparent positive safety signal must be interpreted with caution, given that 73.7% of the included studies failed to report safety data, and none specified laboratory monitoring parameters such as liver or kidney function. While current data do not present clear safety concerns, the lack of comprehensive safety monitoring means that potential risks—such as organ toxicity or long-term medication effects—cannot be ruled out. This gap in safety data represents a major methodological flaw in the current literature, severely limiting its clinical applicability. In terms of secondary outcomes, the combination of CFDTT and conventional pharmacotherapy also showed trends toward improvement. These findings may offer a biological basis for the primary outcomes, suggesting that CFDTT may help optimize ovarian morphology and enhance endometrial receptivity, potentially improving the chances of successful pregnancy. However, these secondary outcomes are similarly constrained by high heterogeneity and low-quality evidence, and their clinical relevance should be confirmed through endpoints like pregnancy rates. As such, this discussion focuses primarily on the interpretation of primary outcomes. To explore potential differences between study populations (e.g., disease duration, age, specific interventions), subgroup analyses were conducted. Sensitivity analyses were also performed to assess the robustness of the results, while publication bias was evaluated using funnel plots and Egger’s test. The quality of evidence was assessed using the GRADEpro tool. Results indicated that CFDTT combined with conventional pharmacotherapy may have advantages in promoting ovulation and improving pregnancy rates. However, all conclusions are tempered by significant limitations in the original studies, including high risk of bias, high heterogeneity, and severe gaps in safety monitoring. In conclusion, this study provides preliminary evidence supporting the use of CFDTT as an adjunctive therapy for infertility associated with PCOS. The findings suggest that combining CFDTT with conventional pharmacotherapy may offer synergistic benefits, while monotherapy with CFDTT appears comparable in efficacy to conventional pharmacotherapy, although the available evidence is limited. It is crucial to highlight that the reliability of the primary outcomes—ovulation rate, pregnancy rate, and safety—underpinning the current evidence base is significantly compromised by methodological shortcomings. Therefore, the conclusions regarding the efficacy and safety of CFDTT should be interpreted with caution, and further validation through large-scale, prospective, and rigorously designed randomized controlled trials is essential. The complex pathogenesis of PCOS-related infertility and the incomplete understanding of CFDTT’s mechanisms remain major challenges for research and clinical application. Based on current evidence, the potential mechanisms of CFDTT can be summarized as follows. C CFDTT appears to improve ovarian function, sex hormone balance, and endometrial receptivity. In PCOS-IR rat models, CFDTT reversed polycystic ovarian changes, reduced ovarian volume, and decreased cystic follicles by activating the IRS1/PI3K/AKT/GLUT4 pathway, enhancing glucose and lipid metabolism [ 36 ]. Clinically, CFDTT reduced ovarian volume, acne scores, menstrual irregularities, and serum reproductive hormones—including FSH, LH, estradiol (E2), testosterone (T), and prolactin (PRL)—while increasing ovulation rates [ 37 ]. CFDTT also improved uterine hemodynamics, reducing endometrial blood flow resistance and enhancing pregnancy outcomes, potentially via estrogen restoration and oatp4a1-mediated sex hormone transport [ 12 , 38 ]. CFDTT modulates glucose and lipid metabolism, as well as insulin resistance, through multiple pathways. In obese PCOS rats, CFDTT corrected hepatic Leptin/AMPK signaling by upregulating Leptin, AMPKα, and CPT1A while inhibiting ACC1, promoting fatty acid oxidation, and ameliorating dyslipidemia [ 39 ]. CFDTT also enhanced insulin sensitivity via the APN/AMPK pathway, improving body fat, phlegm-dampness symptoms, and granulosa cell glucose uptake, which supports oocyte energy supply and embryo development [ 35 ]. Additionally, CFDTT targets the PGC-1α pathway through miR-29a inhibition, promoting mitochondrial biogenesis and energy metabolism in granulosa cells [ 40 ]. Clinically, CFDTT combined with conventional therapy regulates IGF-1, GDF-9, adipokines, and inflammatory factors, optimizing metabolic-endocrine networks to enhance ovulation and pregnancy outcomes [ 5 ]. CFDTT reduces systemic inflammation and oxidative stress. Post-treatment, serum inflammatory markers decrease alongside improvements in gut microbiota diversity, including increased Lactobacillus and Bifidobacterium and reduced Enterobacteriaceae and Bacteroides. Beneficial metabolites (SCFAs, GDCA, TUDCA, CDCA) increase, while harmful bile acids decrease [ 41 ]. CFDTT also inhibits the TLR4/NF-κB p65 pathway in ovarian tissue by downregulating HMGB1, TLR4, and ox-LDL, reducing pro-inflammatory cytokine release (IL-6, TNF-α, CRP) [ 42 , 43 ]. Collectively, these actions modulate the gut microbiota–metabolite–inflammation axis, suppress chronic inflammation, and promote reproductive endocrine recovery. CFDTT maintains ovarian microenvironment homeostasis by regulating autophagy, apoptosis, and cell cycle progression in granulosa cells. In phlegm-dampness PCOS patients, miR-17-5p is downregulated with concurrent ATG7 upregulation, contributing to excessive autophagy. CFDTT restores miR-17-5p expression, inhibits ATG7, and enhances oocyte quality and embryo development [ 44 ]. CFDTT also decreases cleaved caspase-3/9 expression and increases Bcl-2 levels, reducing granulosa cell apoptosis and supporting follicular development [ 45 ]. Additionally, CFDTT promotes PKP3 promoter methylation, suppresses MAPK/ERCC1 signaling, and facilitates granulosa cell proliferation and functional recovery [ 46 ]. In summary, CFDTT exerts therapeutic effects through multi-pathway modulation, improving endocrine, metabolic, inflammatory, and cellular aspects of PCOS pathology. These mechanisms highlight the holistic advantages of traditional Chinese medicine in managing complex reproductive disorders. PCOS, as a highly heterogeneous disorder, requires adherence to the TCM principle of “Syndrome Differentiation and Treatment”, which involves selecting targeted formulas according to the patient’s dominant pathological pattern, such as phlegm-dampness, kidney deficiency, blood stasis, or liver qi stagnation. This study emphasizes CFDTT, a representative formula for PCOS-related infertility with a phlegm-dampness obstruction pattern. Other pathological patterns necessitate distinct formulas (Table  3 ), highlighting TCM’s individualized approach in addressing complex PCOS comorbidities. Table 3 Representative traditional Chinese medicine formulas for treating PCOS-Related infertility according to pattern differentiation Formula Name Core Herbal Components Treated Pattern Targeted Comorbidities & Symptoms Cangfu Daotan Decoction [ 5 , 10 ] Atractylodis Rhizoma, Cyperi Rhizoma, Pinelliae Rhizoma Praeparatum, Poria, Citri Reticulatae Pericarpium Phlegm-Dampness Obstruction Obesity, insulin resistance, ovulation disorders, and polycystic ovarian morphology Dan Zhi Xiaoyao Powder [ 47 – 49 ] Moutan Cortex, Gardeniae Fructus, Angelicae Sinensis Radix, Paeoniae Radix Alba, Bupleuri Radix, Poria, Atractylodis Macrocephalae Rhizoma, Zingiberis Rhizoma Recens, Menthae Haplocalycis Herba Liver Qi Stagnation Transforming into Fire Acne, hyperandrogenemia, and insulin resistance Bushen Huoxue Formula [ 50 , 51 ] Cuscutae Semen, Ligustri Lucidi Fructus, Salviae Miltiorrhizae Radix et Rhizoma, Angelicae Sinensis Radix Kidney Deficiency with Blood Stasis Oligomenorrhea, delayed follicular development, thin endometrium, recurrent miscarriage. Guizhi Fuling Pills [ 51 , 52 ] Cinnamomi Ramulus, Poria, Moutan Cortex, Persicae Semen, Paeoniae Radix Rubra Blood Stasis Pattern Ovulation disorders, insulin resistance Representative traditional Chinese medicine formulas for treating PCOS-Related infertility according to pattern differentiation Effective treatment requires matching the patient’s core syndrome to the appropriate formula, as CFDTT’s efficacy is limited when phlegm-dampness is not the primary pattern. In more complex cases, combining multiple formulas may be required to synergistically target multiple pathological pathways. This study systematically evaluated the efficacy and safety of CFDTT for PCOS-related infertility through meta-analysis and summarized potential mechanisms to provide comprehensive evidence for clinical application. The study quantified CFDTT’s effects on ovulation and pregnancy rates and systematically assessed secondary reproductive outcomes, including ovarian volume, dominant follicle count, and endometrial thickness, offering robust evidence for traditional Chinese medicine interventions. By comparing combination therapy with monotherapy, the analysis distinguished adjunctive versus independent effects of CFDTT, providing refined evidence for clinical decision-making. Subgroup analyses and trim-and-fill correction preserved the robustness of effect estimates, enhancing applicability to heterogeneous real-world patient populations. The research bridged traditional Chinese medicine theory and modern reproductive indicators, validating the biological basis of “phlegm-dampness syndrome” using objective parameters such as ovarian volume, dominant follicle count, and endometrial thickness. Methodologically, the study adhered to PRISMA guidelines, integrated GRADE evidence grading, and employed sensitivity analyses, forming a rigorous quality evaluation loop to ensure cautious and reliable interpretation. However, this study is subject to several methodological and evidence-level limitations that must be considered. First, the high degree of heterogeneity observed in this research represents a significant constraint when interpreting the results. While the meta-analysis revealed substantial pooled effect sizes for CFDTT in improving secondary outcomes such as ovarian volume ( I² = 94%), dominant follicle count ( I² = 78%), and endometrial thickness ( I² = 82%), the marked heterogeneity suggests that the true effect sizes across studies may vary considerably. This variability likely reflects significant inconsistencies in treatment protocols, such as differences in intervention measures, combinations of medications, patient demographics, and methods of outcome measurement—issues that are characteristic of clinical research in traditional Chinese medicine. We attempted to explore potential sources of heterogeneity through subgroup analyses, considering factors such as age, disease duration, and categories of conventional pharmacotherapy used. However, a lack of sufficient reporting on key details, such as the specific composition of “classical versus modified formulas” and the principles of modification, hindered our ability to perform a critical subgroup analysis based on “formula type.” Thus, it is essential to acknowledge that the therapeutic efficacy of CFDTT for PCOS-related infertility is not uniform. Its effectiveness is likely to vary based on the specific formulation, concomitant Western medication regimens, and patient characteristics. Second, the conclusions of this review should be interpreted with caution due to the generally low methodological quality of the included studies. These studies carry a substantial risk of bias, which undermines the credibility of the effect estimates. Our rigorous risk of bias assessment, using Cochrane tools, revealed a concerning pattern: of the 19 RCTs included, only one (5%) employed double-blinding, while the remaining 18 (95%) failed to blind either participants or researchers. Furthermore, all studies (100%) did not report allocation concealment protocols. This widespread methodological deficiency led to the majority of studies being classified as having a “high risk” or “some concern” regarding “deviation from the intervention” and “outcome measurement.” These flaws likely resulted in an overestimation of the efficacy of CFDTT. Consequently, the positive pooled results for ovulation and pregnancy rates observed in this meta-analysis should be interpreted with caution, as they are likely inflated due to the methodological weaknesses in the included studies. This is why the quality of the evidence for the primary outcomes was rated as “low” or “very low” using the GRADE system. Third, the attribution of efficacy remains uncertain. None of the included studies employed placebo controls, and only two RCTs provided direct comparisons with conventional pharmacotherapy. As such, the net therapeutic effect of CFDTT on PCOS-related infertility cannot be definitively determined. Further well-designed head-to-head studies are needed to validate these findings. Fourth, there are ethnic and geographical limitations to consider. The majority of the studies were sourced from Chinese databases, and the genetic background, dietary habits, lifestyle, and environmental factors specific to the Chinese population may influence both the presentation of PCOS and the response to CFDTT treatment. Therefore, the current evidence supporting the efficacy of CFDTT may be most applicable to Chinese populations, and its generalizability to other populations remains uncertain. Fifth, publication bias must be acknowledged. Egger’s tests for ovulation and pregnancy rates ( p  = 0.017; p  = 0.001) and the observed funnel plot asymmetry suggest a potential underreporting of negative or ineffective studies. This introduces the possibility that the observed pooled effect sizes may overestimate the true efficacy of CFDTT. Sixth, there was inadequate protocol transparency across the studies. None of the included trials were pre-registered on platforms such as PROSPERO, and selective reporting of positive outcomes likely contributed to the overestimation of efficacy. Finally, the safety data available in the included studies is insufficient. Seventy-three point 7% of studies failed to report adverse events, and none of the studies outlined predefined laboratory safety monitoring protocols (e.g., liver/kidney function tests, complete blood count, or metabolic indicators). This lack of safety data limits the ability to assess potential risks, such as organ toxicity, and may introduce systematic bias in safety evaluations. Based on the findings and limitations of this study, the following recommendations are provided for future clinical research in TCM: First, future studies should establish consensus and strictly define core diagnostic criteria for TCM syndromes applicable to CFDTT. They must clearly specify the required reporting of formula composition, dosage, decoction methods, treatment duration, and principles for syndrome-specific modifications. Uniform quality standards and specifications should be adopted for both raw herbal materials and ready-to-use granules. Establish rigorous quality standards for herbal materials and implement a traceability system to minimize the impact of herbal material variability on study accuracy. Promote the development of ready-to-use granules and establish strict standards for granule formulations to ensure consistency across studies in herbal preparation. Second, research designs should strictly adhere to internationally recognized clinical research standards, such as the Consolidated Standards of Reporting Trials (CONSORT) statement. Ensure the scientific design and accurate implementation of key elements including randomization methods, blinded allocation, and allocation concealment. Employ computer-generated random sequences, with allocation concealment managed by an independent third party to guarantee randomization and objective integrity throughout the study. Conduct large-scale, multicenter clinical trials to enhance the reliability and generalizability of findings. Third, incorporate placebo controls whenever feasible to eliminate confounding factors like placebo effects and other biases, thereby enabling accurate assessment of the net therapeutic efficacy of TCM. Fourth, increase cross-ethnic and multi-regional clinical studies to incorporate data from diverse racial backgrounds, thereby enhancing the generalizability and external validity of conclusions. Fifth, future research should strictly adhere to the CONSORT (Consolidated Standards of Reporting Trials) statement for reporting, ensuring all predefined outcome measures—including negative results—are fully presented in the final publication. Academic journals, funding agencies, and scholarly organizations should jointly advocate for and establish dedicated channels to encourage researchers to publish their complete findings regardless of outcome. Sixth, standardize trial registration procedures by mandating pre-trial registration and protocol submission for TCM clinical studies, while strengthening post-registration monitoring to prevent selective reporting of outcomes. Seventh, refine the safety evaluation system by systematically and comprehensively recording and reporting all potential adverse events. Particular emphasis should be placed on assessing long-term safety through objective indicators such as liver and kidney function monitoring. Regular follow-ups and dynamic monitoring during the research process should enable the timely identification, management, and documentation of adverse events, thereby enabling a comprehensive assessment of the safety of traditional Chinese medicine formulations.

Polycystic ovary syndrome (PCOS) is one of the most common endocrine-metabolic disorders in women of reproductive age, affecting an estimated 11–13% globally [ 1 ]. It is characterized by hyperandrogenism, chronic anovulation, and polycystic ovarian morphology. Clinically, PCOS presents with menstrual irregularities, hirsutism, obesity, and, most importantly, ovulatory dysfunction, which is a leading cause of female infertility [ 2 ]. PCOS accounts for 6–15% of infertility cases, placing a substantial burden on reproductive health and quality of life [ 3 ]. In recent years, changes in lifestyle and increasing psychological stress have contributed to a rising incidence and earlier onset, amplifying both the reproductive and economic burden on affected women and society. In traditional Chinese medicine (TCM), PCOS is classified under conditions such as “amenorrhea,” “delayed menstruation,” and “infertility.” Its pathogenesis is described as “kidney deficiency as the root, with phlegm and blood stasis as secondary factors.” Insufficient kidney essence impairs follicular development, while phlegm-dampness accumulation obstructs the uterus, disrupting qi and blood flow and impairing ovulation. Liver qi stagnation further exacerbates phlegm-dampness and blood stasis, creating a self-perpetuating cycle of dysfunction [ 4 , 5 ]. Modern medical management of PCOS-related infertility focuses on endocrine regulation, ovulation induction, and assisted reproductive technologies. Agents such as clomiphene, metformin, and Diane-35 show benefits but are limited by adverse effects, resistance, or suboptimal efficacy [ 6 , 7 ]. Surgical options, including ultrasound-guided follicular aspiration and laparoscopic ovarian drilling, carry risks of complications such as ovarian adhesions or hypoplasia‌ [ 8 ]. In vitro fertilization (IVF) can improve pregnancy outcomes but is costly, technically demanding, and associated with ethical and safety concerns‌ [ 6 ]. These limitations highlight the need for safer, more effective, and less invasive treatment options. ‌ TCM offers a holistic approach, emphasizing multi-target and multi-pathway regulation. CFDTT, recorded in Ye Tianshi’s Secret Formulas for Gynecological Diagnosis and Treatment of the Qing Dynasty, is a classic prescription for infertility due to phlegm-dampness obstruction. It is formulated on the principle of “drying dampness, resolving phlegm, and regulating the Chong and Ren meridians.” Atractylodis Rhizoma serves as the principal herb to dry dampness and strengthen the spleen, Cyperi Rhizoma regulates qi and alleviates stagnation, Pinelliae Rhizoma and Arisaema cum Bile transform phlegm and disperse nodules, Pericarpium Citri Reticulatae and Poria drain dampness, and Aurantii Fructus regulates qi and unblocks meridians. Together, these actions aim to clear phlegm, restore circulation, and enhance uterine function‌ [ 9 ]. Accumulating clinical evidence suggests that CFDTT may improve ovulation rates (OR = 1.89 [1.32–2.71]), regulate endocrine profiles, and enhance pregnancy outcomes, with a generally favorable safety profile‌ [ 5 , 10 – 12 ]. However, most existing studies are limited by small sample sizes, single-center designs, and methodological weaknesses. High-quality systematic reviews and meta-analyses are lacking, and the efficacy and safety of CFDTT have not yet been comprehensively assessed. Therefore, this study seeks to conduct a systematic review and meta-analysis to rigorously evaluate the efficacy and safety of CFDTT as an adjunctive treatment for infertility associated with PCOS. The findings will provide evidence-based support for the rational clinical application of this formulation, thereby contributing to the integration of traditional Chinese medicine in the management of PCOS-related infertility.

Supplementary Material 1 Supplementary Material 1