The research methodology utilized exclusively pre-existing scholarly data, thereby eliminating the requirement for institutional review board authorization or individual participant agreements.

Two investigators working independently executed a comprehensive literature retrieval strategy, examining four major international databases (including PubMed, Embase, Web of Science, and ClinicalTrials.gov) alongside three prominent Chinese academic repositories (CKNI, Wanfang, and CQVIP), with the search window extending through March 31, 2025. Search for randomized controlled trials (RCTs) on the use of Niaoduqing Granules in managing uremic pruritus among hemodialysis patients. The search strategy was consisted of free word and MeSH terms, which included: Niaoduqing Granules, uremic pruritus and hemodialysis. Manual searches were conducted through the reference lists of included studies and existing meta-analyses to identify potentially eligible randomized controlled trials (RCTs).

Studies were incorporated into the analysis according to the following criteria in line with the PICOS framework: (I) Participants: patients presenting with hemodialysis-dependent uremic pruritus; (II) Interventions: Hemodialysis patients with uremic pruritus treated with Niaoduqing Granules versus those not treated with Niaoduqing Granules; (III) Comparisons: comparison of treatment safety and pruritus improvement between hemodialysis patients receiving versus not receiving Niaoduqing Granules; (IV) Outcome measures: studies needed to report one or more of the following endpoints—visual analogue scale, 5D-itch scale, effective rate, adverse drug reaction, serum calcium, serum parathyroid hormone, serum phosphorus, serum intact parathyroid hormone, serum creatinine, blood urea nitrogen, uric acid, β2-microglobulin, high-sensitivity c-reactive protein, tumor necrosis factor-α, and interleukin-6; (V) Study design: eligible studies were limited to official, reported, and full-text randomized controlled trials (RCTs).

The exclusion criteria were as follows: (I) studies that did not meet the inclusion criteria, (II) articles in the form of editorials, conference proceedings, commentary articles, reviews, clinical conference reports, and abstracts, (III) studies involving animals, and (IV) duplicate reports from the same database.

Two independent reviewers performed study selection and data extraction. Discrepancies were resolved through third-author consultation and consensus. Extracted data from eligible studies included: Two independent reviewers performed study selection and data extraction. Discrepancies were resolved through third-author consultation and consensus. Extracted data from eligible studies included: Furthermore, the outcome data were also collected, which comprised: the assessment of visual analogue scale, 5D-itch scale, effective rate, adverse drug reaction, additionally, changes in serum calcium, serum parathyroid hormone, serum phosphorus, serum intact parathyroid hormone, serum creatinine, blood urea nitrogen, uric acid, β2-microglobulin, high-sensitivity c-reactive protein, tumor necrosis factor-α, and interleukin-6, were analyzed following Niaoduqing Granules administration.

We used the Cochrane Collaboration tool to evaluate the quality of all included RCTs. More specifically, this study used a ‘risk of bias’ table to show the risk of bias in each trial, including factors like allocation concealment, blinding of outcome assessors, incomplete outcome data, selective outcome reporting, random sequence generation, and other possible sources of bias. Two reviewers independently assessed bias risk (high/unclear/low), with disagreements resolved by consensus or third reviewer involvement. Most RCTs demonstrated moderate quality.

Statistical analysis utilized Stata, assessing heterogeneity with I² statistic (<50% threshold). Random-effects meta-analysis followed Cochrane guidelines. For continuous outcomes, such as visual analogue scale, 5D-itch scale, serum calcium, serum parathyroid hormone, serum phosphorus, serum intact parathyroid hormone, serum creatinine, blood urea nitrogen, uric acid, β2-microglobulin, high-sensitivity c-reactive protein, tumor necrosis factor-α, and interleukin-6, the results were presented as weighted mean differences (WMD) with 95% confidence intervals (CIs). Categorical outcomes, including the overall effectiveness rate and adverse drug reaction, were assessed using risk ratio (RR) with 95% CIs. Statistical significance was determined when the p-value was less than 0.05.

Initial database searches and supplementary sources yielded 208 potentially relevant articles. Following the import of search results into Endnote software, 142 duplicates were systematically removed. Subsequent title/abstract screening excluded 44 studies deemed irrelevant to the research objectives. The remaining 22 articles underwent full-text assessment, during which rigorous application of predefined inclusion/exclusion criteria and evaluation of data completeness were implemented. Ultimately, 14 randomized controlled trials (RCTs) met the inclusion criteria and were included in the meta-analysis (12–25). The comprehensive selection workflow is presented in Figure 1.

The characteristics and demographic data of the patients included in the 14 trials involving 1,232 patients conducted in China are presented in Table 1. All selected studies were RCTs published between 2012 and 2023. Among the 1,232 subjects, all were uremic patients undergoing hemodialysis, all of whom exhibited uremic pruritus of differing severity levels. Of these, 618 patients received Niaoduqing Granule, while 614 did not. Across the 14 included studies, 11 used 5 g orally four times daily (12, 13, 15, 16, 19, 20, 22–25). The remaining 3 studies used alternative regimens: one with 5 g/100 mL twice daily (17), one with 2.5 g/100 mL twice daily (14), and one via enema (15 g/150 mL twice daily) (21). The duration of Niaoduqing Granule treatment varied from 4 weeks to 72 weeks. Of these studies, 10 studies provided information on effective rate (13, 15, 16, 18–24), and 3 studies provided information on adverse drug reactions (12, 20, 25). Of note, there are 10 RCTs that reported the assessment of the severity of pruritus (13, 15–17, 19–24). Further details are provided in Table 1.

Based on the extracted data (Table 1), subgroup analyses were conducted concerning dosage and treatment duration. The majority of the included randomized controlled trials employed a consistent regimen of oral Uremic Clearance Granules at 5 grams, four times daily. Analyses focusing on this standard dosage demonstrated its superior effectiveness in improving pruritus compared to control groups. In terms of treatment duration, which varied across studies, the most robust evidence supports a medium-term course of approximately 12 weeks. While data on long-term efficacy (>3 months) are promising, they are less abundant. Consequently, the most substantiated regimen appears to be 5 g orally, four times daily for 12 weeks. However, definitive comparisons with alternative dosages or administration routes are precluded by the limited number of relevant studies. Therefore, future research should include to comparing different regimens and validate long-term sustained benefits.

Two independent reviewers assessed the risk of bias and reporting quality of the included studies through the Cochrane Collaboration’s evaluation tool, categorizing each criterion as low, unclear, or high risk. Disagreements between reviewers will first be addressed through discussion, with unresolved issues escalated to a third reviewer for arbitration. Each study will be scrutinized across seven domains: randomization methods, allocation concealment, blinding of participants and researchers, blinding of outcome assessors, handling of incomplete data, selective reporting, and other biases. Most studies exhibited low risks in randomization, outcome assessor blinding, data completeness, and reporting transparency, while challenges persisted in allocation concealment and participant blinding. Overall, as summarized in Table 2, most of the included RCTs were of moderate quality.

A total of 14 RCTs involving 1,232 patients with uremic pruritus, were included. Detailed results are presented in Table 3 and summarized as follows.

Data from eight studies indicated that the Niaoduqing Granule group had significantly lower visual analogue scales than the control group (WMD = −2.151; 95% Cis = −2.372 to −1.930; p < 0.001; I² = 64.2%; Figure 2). Data from two studies indicated that the Niaoduqing Granule group had significantly lower 5D-itch scale than the control group (WMD = −7.642; 95% Cis = −14.042 to −1.243; p = 0.019; I² = 99.2%; Figure 2). Notably, a marked attenuation of uremic pruritus severity was achieved in patients following Niaoduqing Granule.

Ten studies reported the effective rate, with 421 patients in the experimental groups and 415 patients in the control groups. There was no significant heterogeneity (p < 0.001; I² = 0.0%), meanwhile, we applied a random-effects model. The results revealed a statistically significant difference between the two groups (RR = 1.249; 95% CIs = 1.169–1.335; p < 0.001; Figure 3). Adverse drug reactions were assessed in three trials, showing no significant difference between the Niaoduqing Granule group and control groups (RR = 1.303; 95% CIs = 0.589–2.884; p = 0.514; I² = 0.0%; Figure 3). Niaoduqing Granules treatment demonstrated significant therapeutic efficacy enhancement in patients, with no concomitant increase in adverse drug reactions, thereby validating the regimen’s safety profile.

Data from eight studies indicated that had no significant difference in serum calcium levels was observed between the two groups (WMD = −0.041; 95% CIs = −0.577 to 0.495; p = 0.881; I² = 99.78%; Figure 4). Five studies find significant difference serum parathyroid hormone levels was observed between the two groups (WMD = −93.026; 95% CIs = −141.908 to −44.144; p < 0.001; Figure 4).

Nine studies found a statistically significant difference in serum phosphorus levels between the two groups (WMD = −0.290; 95% CIs = −0.427 to −0.154; p < 0.001; Figure 5). Four analyses do not find a significant difference between the two groups Serum intact parathyroid hormone levels, whereas, a high heterogeneity presents in these studies (WMD = −93.383; 95% CIs = −238.012 to 51.246; p = 0.206; I² = 99.2%; Figure 5).

A pooled analysis of 578 patients from six studies showed a reduction in Serum creatinine levels (WMD = −198.788) and blood urea nitrogen levels (WMD = −7.173), with p values indicating statistical significance (p = 0.002 and p < 0.001, respectively). However, due to the extremely high heterogeneity (I² > 97%), these pooled estimates should be interpreted with extreme caution. The substantial variability among studies precludes drawing reliable conclusions about the effect (Figure 6).

Four studies evaluated serum Uric acid levels, upon pooling the results, no significant difference in uric acid levels was observed between the two groups (WMD = −24.518; 95% CIs = −64.240 to 15.204; p = 0.226; Figure 7). However, the results showed that the levels of β2-microglobulin in the two groups differed statistically significantly. (WMD = −9.270; 95% CIs = −11.613 to −6.927; p < 0.001; Figure 7).

Data from seven studies indicated that high-sensitivity c-reactive protein, tumor necrosis factor-α and interleukin-6 levels all demonstrated statistically significant difference between the two groups. As summarized in Figure 8 (High-sensitivity C-reactive protein: WMD = −1.801; 95% CIs = −2.539 to −1.063; p < 0.001; Tumor necrosis factor-α: WMD = −15.234; 95% CIs = −20.002 to −10.465; p < 0.001; Interleukin-6: WMD = −6.133; 95% CIs = −7.422 to −4.844; p < 0.001), whereas, there was no significant heterogeneityininterleukin-6 levels and tumor necrosis factor-α levels. Demonstrates efficacy in attenuating inflammatory burden among hemodialysis patients.